Featured Story
Nokia alert to risks as it puts AI agents into mobile core #
A new service launched by Nokia this week gives operators a preview of 'AI-native' features baked into its latest mobile core products.
Outside urban hotspots, the 6GHz and even higher spectrum auctioned off for 6G services is unlikely to hold much value. Signals turn limp and lethargic as they ascend the frequency ladder, struggling to overcome the flimsiest barriers or travel very far. Vendors think building even more advanced radios is the answer. But many operators would prefer to refarm the lowband spectrum originally assigned to older generations.
All that makes MRSS, a 6G technology designed for this purpose, look far more important than DSS, a 5G predecessor with a bad reputation. "We need MRSS to be a lot better than that," said Luke Ibbetson, the head of research and development for Vodafone. "Otherwise, it's going to be challenging to move into 6G."
Standing for multi-RAT (radio access technology) spectrum sharing, MRSS should allow an operator to divide in-use frequencies between current 5G and future 6G services. The danger is that operators encounter the same problems they faced with DSS (dynamic spectrum sharing), which similarly split resources between 4G and 5G.
"We don't want there to be an impact on the 5G subscribers when we start sharing the spectrum between 5G and 6G," said Ibbetson. "This is where the dynamic spectrum sharing stuff was really not that well implemented."
Nick Read, Vodafone's previous CEO, was scathing about DSS way back in late 2020, when he offered some harsh criticism of the technology on a call with equity analysts. "Some operators are taking DSS, which is effectively giving you a 5G symbol but 4G performance," he said at the time.
If MRSS can avoid acquiring the same reputation, it could establish itself as the mainstream technology for implementing 6G. But this could produce another problem by tying operators to the same vendors they used in 5G. For operators like Vodafone, the MRSS objective is not just to refarm spectrum but also about reusing equipment. "If you look at the fundamental rationale for this, it's to avoid the need to go and replace radio infrastructure," said Ibbetson. Deployed that way, MRSS would not support the introduction into the network of a new vendor with 6G's launch. The desire to reuse hardware makes any talk of multivendor MRSS sound "moot" to Ibbetson. "If you're putting a brand new radio into a brand new frequency band, then nothing in the way 6G has been specified would prevent you from using a new vendor," he said. "But then, by definition, that new vendor doesn't have any 5G gear on the site."
Yet there is an alternative MRSS scenario that parts of the industry have examined. Described in a white paper by Nokia several years ago, it would still entail the sharing of a single radio unit (RU). But it would also allow an operator to introduce a new 6G vendor for the distributed unit (DU), the hardware box (a server or appliance) that handles signal processing, and run it alongside the existing 5G DU.
That set-up would, however, require the standardization of a new interface in the network to "facilitate coordination across packet schedulers," as Nokia described it. No such new interface had been developed for DSS, and the Finnish company did not sound enthusiastic about the idea of one that would link DUs from different vendors. "Avoiding new interfaces is preferable, given the associated standardization, implementation, testing, and integration costs," it said.
Those hoping for progress and a change of view since then are likely to feel disappointed. Approached by Light Reading last week, rival Ericsson was similarly dismissive, insisting multivendor 6G does not require a new interface between DUs. It is already possible, said Ericsson, thanks to earlier work on developing an open fronthaul (O-FH) interface that joins a DU from one vendor to an RU from another. That was carried out years ago as part of the industry's open RAN (O-RAN) mission.
"The alternative scenario – separate 5G and 6G systems, potentially from different vendors, sharing the same radio unit – would be complex to realize in practise," said an Ericsson spokesperson by email. "It would require very close coordination across radio, baseband and software layers, O-FH support across the relevant 5G and 6G nodes and could involve performance trade-offs compared with a more integrated, single-vendor MRSS implementation."
Whether implementation would be complex or not, the attitude of both Ericsson and Nokia is unlikely to please anyone who wants to see multivendor MRSS made possible. "Vendors are reluctant to define a new real-time interface between DUs from two different vendors," said Alok Tripathi, an independent analyst and former technology expert for China's Huawei who covers RAN matters in some detail on LinkedIn.
As Tripathi pointed out during an exchange of messages with Light Reading, parties interested in the concept of multivendor MRSS would appear to include NTT Docomo, Japan's biggest telco. Its representatives told a 3GPP workshop last year that operators should have the "freedom" to do what they call "inter-vendor MRSS."
Any performance gap between intra- and inter-vendor MRSS would mean "such a freedom will be significantly limited," said NTT Docomo in a presentation (PDF download). Among other questions raised by the operator was one about the "level of granularity for scheduling coordination" that could be expected. "How to reduce the performance gap between intra- and inter-vendor MRSS?" it asked.
If multivendor MRSS remains unviable, the industry still has the O-FH ability to pair RUs and DUs from different vendors, according to Ericsson. "We expect open fronthaul (O-FH) to remain the primary standardized route for multivendor RAN deployment in 6G, enabling a radio unit from one vendor to interoperate with a distributed unit from another," said the company in its email. "O-FH already underpins multivendor deployment in 5G through O-RAN, and we expect the interface to evolve to support 6G from the outset, including 5G/6G multi-RAT scenarios," the vendor added. "Ericsson will support O-FH in its 6G portfolio."
As positive as all that may sound, this species of multivendor RAN has become harder to spot than mating pandas. For a mixture of commercial and technical reasons, most operators prefer to source all the equipment at a given macro site from the same vendor's system.
"Multivendor RAN outlook remains subdued," said Stefan Pongratz, an analyst with Dell'Oro, in a recent LinkedIn post about the market situation. "Broader adoption of open fronthaul is not expected to significantly alter RAN supplier concentration, as most deployments will continue to use radios and basebands from the same supplier."
Vendors have also been vanishing, and more of them could disappear by the time 6G is launched in the 2030 timeframe. That is partly a consequence of the weak appetite for multivendor RAN. What Dell'Oro calls "single-vendor open RAN" has naturally meant that telcos continue to buy O-RAN-compliant products from the same big suppliers. Without revenues, challengers have lacked the funds they need to invest in product development and compete.
Even the big vendors are now suffering as operators slash RAN investment, disappointed by the returns they have seen years after the initial 5G splurge. Spending on RAN products fell from $45 billion in 2022 to $35 billion last year, said Omdia, a Light Reading sister company. It forecasts relatively unchanged revenues this year.
The difficult environment has prompted Ericsson and Nokia to cut tens of thousands of jobs from their combined headcount this decade as profit margins are crushed. Pressure seems unlikely to ease. Dell'Oro now expects cumulative RAN revenues to be 10% to 20% lower in 6G than they were in 5G for the first six years after launch. If MRSS takes off, and lets operators sweat their 5G hardware for years, suppliers could be in for a painful time.
Read more about: