Standard, open interfaces at every level of the network and device are essential if 5G is to achieve its goal of enabling massive but affordable density – billions rather than millions of connected devices, up to 10 times more base stations, and ubiquitous coverage indoors and out. That requires networks and devices that are far cheaper, and far easier to deploy and manage, than traditional cellular systems.
The open interfaces which will stimulate a more competitive, innovative device ecosystem, and drive the development of low cost handsets, are starting to emerge in multiple layers, including the APIs that enable a common base of applications and tools across smartphones.
The latest contribution is focused on the other end of the stack, with the formation of the Open RF Association (OpenRF), which aims to reduce time, cost and complexity for a wide range of companies to build a smartphone or other mobile device. That would open up the relatively small clique of companies which have a major presence in smartphones, and so drive additional business for the RF industry, argue the founders of OpenRF – led by Broadcom, Intel, MediaTek, Murata, Qorvo and Samsung.
Their plan to develop a common framework for hardware and software, across RF front end and modem platforms, has to be seen as a defensive move against Qualcomm, which has been steadily expanding its activities in the RF front end (RFFE), and has targeted this as its next big growth area in the device market. The RFFE is the group of components which transmits and receives signals between the radio transceiver and the antennas, and includes signal filters, power amplifiers and antenna switches.
Success for Qualcomm – leveraging tight integration with its smartphone processors, modems and systems-on-chip (SoCs) – would not only pressurize RF specialists like Qorvo, but also give it a competitive advantage against other smartphone SoC providers, like MediaTek or Samsung.
This summer, Qualcomm said it aimed to be the market leader in RFFE by the end of this year and to seize more than 20% of the market by the end of 2022, claiming its modem-to-RF portfolio is a key differentiator.
Cristiano Amon, president of Qualcomm, has repeatedly used the prospect of 5G to pursue a familiar Qualcomm theme – the need to integrate as many components as tightly as possible, in order to reduce cost, size and power consumption of a device, and maximize performance. In 5G, it will very difficult to build a radio platform which does not have the front end fully integrated because of the number of bands that will be supported, he argues.
OpenRF said its proposed standards would benefit chip and device makers with R&D savings, faster time to market, greater simplicity of design for OEMs, improved economies of scale and reduced inventory costs and overall risk. The group published a white paper, commissioned from Mobile Experts, to analyze the potential upsides. Chief analyst Joe Madden said: “The number of smartphones sold each year has stopped growing, but the level of complexity for the RF chain continues to grow and has become truly mind-boggling.”
This complexity is heavily tied into the growing number of spectrum bands and combinations that will be supported by 5G. The number of RF band configurations could from about 1,000 for 4G to over 10,000 for 5G, and that leads to yet more permutations, such as different types of filters and MIMO antennas to handle different frequency ranges effectively. Carrier aggregation can further complicate the choices, and if an OEM has to start from scratch to design a device every time it wants to support a different set of RF options, the cost escalates quickly.
The new group has set up five working groups to produce its initial specifications, APIs and frameworks. The first group will create a register map framework, defining “a set of core chipset and RF front end features and interfaces that will enable interoperability across 5G basebands while allowing innovation across vendors”.
The other four are:
- The software API working group – targeted to develop “a common hardware abstraction layer enhancing the transceiver/modem and RFFE interface”
- The RFFE/RFIC group – it will specify a set of common hardware interface features across chipsets
- The RF power management group
- The compliance program.
Mobile Experts says the smartphone ecosystem could save $900m a year, collectively, in R&D costs by standardizing common elements that do not provide competitive differentiation. Currently, the RFFE suppliers alone spend about $1.5bn a year on R&D. Smartphone vendors would benefit the most, saving an estimated $550m, says the white paper, while modem suppliers could save about $220m, and RFFE specialists about $130m.
“OpenRF provides the framework to streamline and condense the OEM’s design cycle, from inception to product launch. This is a critical step to promote a pro-competitive environment, allowing OEMs to freely choose solutions based on performance, size and cost,” said David Archbold, VP of marketing at Broadcom’s Wireless Semiconductor Division, in a statement.
While not ostensibly sparked by Qualcomm’s ambitions in the RFFE world, one goal is clearly to dilute the US company’s integration advantages by enabling pre-integration, at low cost. And the new group was not shy of criticizing the smartphone chip giant. “Qualcomm doesn’t participate because Qualcomm’s strategy is to lock the OEMs into using Qualcomm RF modules with Qualcomm power supplies and Qualcomm modems,” Madden told LightReading. “Open interfaces run counter to Qualcomm’s basic strategy.”
Qualcomm signalled its intentions clearly a year ago when it took full control of RF360 Holdings, which it set up in 2017 as a $3bn joint venture with TDK of Japan. In buying out its partner, the chip vendor said it was demonstrating how successful RF360 had been so far. Full control gives it enhanced ability to fill gaps in its end-to-end 5G device chip platform, and if it can provide as many components as possible, this can offset the pressures on the addressable market for its core processor/modem offerings.
The TDK venture helped Qualcomm to build on, and expand, the work it had previously done to move into the RF front end (RFFE) space and so address a larger number of elements in a cellular device. The joint venture was named after its first major product in that area, also called RF360. At the time of its formation, CEO Steve Mollenkopf said the JV would “transform us from a strong niche player to a formidable challenger with all the necessary IP blocks” in RF front ends.
In particular, the TDK relationship brought Qualcomm a range of filters which have enabled it to support a wide range of spectrum bands, from sub-GHz to millimeter wave, and to offer an end-to-end solution, integrated from modem to antenna. This is the Snapdragon 5G Modem-RF System, which integrates power amplifiers, filters, multiplexers, antenna tuning, low noise amplifiers, switching and envelope tracking.
Qualcomm said the acquisition of TDK’s share was the final step in bringing more than 20 years of expertise in RFFE filtering inhouse, with products spanning all the main RFFE filter technologies including surface acoustic wave (SAW), temperature-compensated SAW (TC-SAW), thin film SAW and bulk acoustic wave (BAW). The end result is a fully integrated solution combining these filters with Qualcomm components including power amplifiers, envelope tracking and switches (the last of these categories enhanced in 2015 by the acquisition of Antenna Tuning). At the time of the JV being announced, TDK was shipping over 25m filter functions a day, including design wins with all the top tier smartphone makers, and saw demand growing.
In 2013, Qualcomm took the gamble of playing for more of the real estate in a device by offering the RF as well as the processor and modem. That has paid off over the past year or so, and at the 2018 Consumer Electronics Show, Qualcomm announced RFFE deals with Google, HTC, LG, Samsung and Sony Mobile.