Your browser is not supported. Please update it.

5 March 2019

Ericsson buys Kathrein’s antennas as vendors seek control of key elements

As operators see open interfaces and mix-and-match networks tantalizingly in sight for 5G RANs, the major vendors are consolidating their hold on as many elements of the network as they can, as illustrated by Ericsson’s decision to buy the mobile business of its long-time antenna supplier Kathrein.

The growth and power in the RAN may have shifted to software, but there are, of course, significant physical elements that cannot be commoditized and which still offer a vendor significant revenue and differentiation. These are, notably, the antenna – increasingly complex in 5G with Massive MIMO – and the baseband processor. While arguably the latter could be a standard x86 (or ARM) CPU, in reality the heavy demands of running a RAN baseband in software requires a collection of surrounding accelerators and various types of chips, integrated and optimized in a way that is far from commoditized.

There are two strategies for large OEMs – rely on specialist merchant suppliers for antennas and baseband chips, or go back to the old days of 2G and 3G and design these technologies inhouse, or with heavily controlled, tame third parties. The pendulum is swinging to the inhouse approach, with Nokia creating its ReefShark processors, Huawei making its own antennas and RAN silicon, and now Ericsson acquiring Kathrein.

CEO Börje Ekholm said the German firm’s mobile antenna and filters business would be incorporated into Ericsson’s Networks division, bringing a dowry of 4,000 staff and about €270m (US$306m) a year in revenues, excluding sales to Ericsson itself. But most importantly, it brings full control of the direction of a key technology enabler, and keeps this out of the hands of competitors. The fewer high quality base station antenna providers there are, the higher the barriers to entry for any new challengers to Ericsson, though the deal should boost surviving independents like CommScope and Fractal.

“Antenna design will be increasingly important to accommodate new frequencies and multiple technologies,” he said during his MWC keynote. “At the same time, radios and antennas are being consolidated and integrated to optimize site space and overall network performance. More powerful 5G technologies will depend even more on antenna technologies for performance, coverage and capacity.”

His views were echoed by the head of the Networks business, Fredrik Jejdling, who said: “Strengthening our inhouse antenna competence is another important step in our networks portfolio strategy. The acquisition of Kathrein’s antenna and filters business will expand our capabilities and competences in the advanced active and passive antenna domain further. With the additional focus on the antenna and filter business led by Kathrein professionals, we will broaden our offering to further optimize site space, which is vital for the introduction of 5G.”

Kathrein was an early mover in active antennas, having acquired pioneer Ubidyne in 2013, and has a strong patent portfolio related to that according to Ericsson (last year it settled a patent dispute with CommScope). However, this technology took far longer to go mainstream than initially hoped, and is really being driven by the 5G move to higher order MIMO for dense capacity and high frequency spectrum. Combined with 3D beamforming and self-optimizing network (SON) tools, it can greatly reduce the size and spectral efficiency of a macrosite.

The company has engaged in 5G trials with several operators directly, including Orange. Its remaining businesses focus on satellite equipment, radio broadcasting and emergency services communications.

While this acquisition is focused on the challenges of boosting performance, while reducing site space, for 5G macrocells, there are equally daunting engineering issues for indoor cellular. There are more independent players in this field, including DAS (distributed antenna system) and small cell providers like Corning and CommScope. However, Ericsson wants to assert more control over the technology here too, and at MWC, it unveiled its Radio Stripe.

Currently in the labs, and possibly five or more years from commercial reality, this antenna looks like a strip of sticky tape and is designed to fit into tiny form factors for ubiquitous indoor deployment, going a step beyond Ericsson’s Radio Dot indoor system in terms of coverage, deployability and cost/power efficiency.

“We believe the next step will be to distribute antennas everywhere and we’re already pushing antenna development into the size of a matchbook,” said Ekholm. “But in our labs, we’re also looking to create antennas so small and so flexible that we can integrate them with adhesive tape. We call them radio stripes, and they could actually provide truly ubiquitous connectivity in dense cities.”

He added: “We believe this is going to revolutionize the future of antennas”, especially for environments like stadiums, factories and warehouses. Ericsson’s Jan Hederen said the stripes would be able to be camouflaged as wallpaper or curtains. “This is a way to offer distributed serialized massive MIMO with beamforming,” he said. “It’s the super edge.”