Advances in RF front-ends made 5G phones possible: Lam

Wayne Lam

A decade ago, a fundamental change began in the radio frequency (RF) designs of cellular devices that would pave the way for the current sophistication of the RF front-end — the section of a smartphone between the modem and the antenna. The world was going through a momentous transition from multiple, fractured 3G global cellular technologies to a unified 4G wireless standard that would become known as Long-Term Evolution, or LTE. It was the first step in a long journey of improving cellular communication to make optimal use of radio spectrum.

The advent of 4G LTE called for a new set of wireless technologies to harness available cellular radio spectrum and make the most efficient use of disaggregated spectrum holdings. To accomplish this, RF engineers fundamentally reworked the wireless signaling and transmission architecture in the RF front-end, adding major capabilities such as carrier aggregation, higher-order modulation and multiple-input multiple-output (MIMO) antennas. By combining more and more radio spectrum to improve overall wireless connection and performance, RF designs in mobile devices became considerably more complex.

This is the first of a three-part series that aims to help make sense of the enigma that is RF front-end design and the technologies that make 5G smartphones possible. It explores why RF front-end designs in phones have become so complex, and reviews RF front-end architecture, illustrating how the chipset industry has managed this complexity and continued to improve the user experience.

The RF front-end enigma

Before 4G, cellular radio design was a pretty simple affair. Traditional RF front-ends supported only a handful of different radio frequencies, which, in turn, only needed a small number of RF components and antennas to support the downlink and uplink functionalities. But as the industry embarked on the long-term evolution with 4G, it was evident that RF front-end designs had to be able to scale quickly to match the rise in radio frequencies becoming available around the globe for use in cellular applications.

Today, it's not unusual to see RF front-end designs that support well over 20 frequency bands as well as multiple antennas in 4G and 5G phones. Initial complications in 4G created a geometric growth in the RF front-end section, as compared with 3G. As the industry transitions to 5G, the RF design challenge has compounded, presenting device-makers with the daunting task of arresting that exponential growth in RF complexity. In fact, our research shows that the RF front-end saw the biggest rise in bill of materials (BOM) cost of any section in smartphones. Whereas other functional areas experienced only modest increases in cost and complexity, the RF-front end increased in both cost and complications.

The latest 5G RF front-end designs have to cope with even more network requirements to support the new, wider bandwidth of 5G frequencies, as well as a growing set of LTE frequency bands as the majority of 5G network rollouts at launch used the non-standalone implementation of 5G. This means that two different radios are simultaneously active, one 5G and the other 4G, with unique and separate RF chains.

The 5G standard also helped to usher in new, previously unused spectrum starting at 24 GHz, commonly referred to as millimeter wave spectrum. Upward of 1 GHz of spectrum is available in these bands, enabling peak wireless data speeds more than 7 Gbps.

But to achieve these speeds, device designs sacrifice signal coverage. Reception and propagation of millimeter wave 5G are more difficult in real-world applications, forcing the use of novel radio techniques such as beam-forming and beam-steering to produce a usable millimeter wave 5G connection.

As 5G increases the number of radios and RF components in phones, it has become more difficult to create a functional RF front-end that supports three generations of cellular radio technologies.

The table below illustrates increases in the number of bands supported in 3G, 4G and 5G networks that have led to the runaway growth in RF complexity. Comparing the first-generation designs of the Samsung Galaxy series over the past 11 years, we can see that each new network generation brought more supporting RF bands, increasing the requirements of the RF front-end to keep up with added bandwidth. This meant that handset-makers needed to invest heavily in designing more capable RF front-ends, containing cost inflations and working with limited space on their devices.

CCS graphic 4
Number of radio bands supported by three devices with each new network generation.

All the while, these RF complications are obscured in product literature and RF design challenges hidden from consumers, because the complexities would muddy the product messaging. Smartphone users don't need or want to understand this increasing complexity; they expect their phones to work anywhere regardless of where they live.

Why does it have to be so complicated?

To achieve an order-of-magnitude improvements in network and user experience from generation to generation, more radio spectrum is brought to bear. But the inconvenient truth is that RF front-end design doesn't scale effectively with increased demands on the phone radio. Because spectrum is a scarce resource, governments have to ration RF spectrum, creating a situation where most cellular networks today fail to meet the expected needs of 5G.

Millimeter wave technology is one solution, but it doesn't travel so well, so RF designers can't rely solely on it to solve the spectrum crunch. They need to address the widest set of RF support and architect the most capable cellular radio design ever introduced into a consumer device.

From sub-6 GHz to millimeter wave, all available spectrum must be used and supported in modern radio and antenna design. And because of inconsistent spectrum holdings, both frequency division duplex and time division duplex capabilities have to be combined in a single RF front-end design. Additionally, carrier aggregation, which helps to increase the virtual bandwidth pipeline by bonding spectra in different frequencies, compounds the requirements and complexity of the RF front-end.

Furthermore, the evolving capabilities of wireless local area networks and Wi-Fi add another layer of complication, as cellular and Wi-Fi signals have to be kept apart or otherwise risk massive RF interference. The latest Wi-Fi 6E standard adds 6 GHz spectrum to the mix. So, RF front-ends must have advanced filtering technologies to avoid overlapping RF signals.

Smartphone RF front-end designs have become so complicated because of the exponential growth of RF requirements in 5G. Consumers today expect manufacturers to solve all these RF challenges but don't fully understand how we got to this level of complication. In other words, there's no escape from the complexity of the RF front-end if the promise of 5G is to be realized.

RF front-end designs in existing 5G phones

The RF challenges today manifest themselves in many ways. The number of physical antennas in leading flagship devices rose significantly from the typical three or four to upward of 12 — excluding millimeter wave antenna modules. The following image highlights the growing number of antennas required in modern designs, in addition to the need to incorporate millimeter wave antennas.

CCS graphic 3
Left: a set of sub-6 GHz antennas on flexible substrate corresponding to their placement on the mid-frame of the Realme GT. Right: array of antennas found in the Samsung Galaxy A42 including two millimeter-wave antennas in package.

Moving from 4G to 5G, more antennas are needed to support 4x4 MIMO antennas and to cover a wide range of sub-6 GHz frequencies from 600 MHz to 7 GHz. The use of antenna tuners has helped to repurpose existing antennas for multiple frequencies, reducing the number of physical antennas. Although this is helpful, the non-standalone implementation of 5G calls for tens of thousands of carrier aggregation combinations and special provisions for millimeter wave spectrum demand multiple radio chains. This increases the component count, cost and intricacies of the RF-front end. It's hard to escape the growing and evolving needs of 5G RF front-end designs in 5G smartphones; the best the industry can do is manage that complexity.

As the RF front-end in smartphones contends with exponential growth in RF requirements, it's also being pushed to take less space in the device. In other words, it's stuck between a rock and hard place, trying to balance these two conflicting needs.

So how does the 5G RF front-end remain compact while becoming more capable? The answer is through electronic integration. Below is an image of a front-end module that works as an RF receive path or chain for incoming signals. Modules help to reduce the size of electronic components on printed circuit boards, which, in 5G smartphones, are very limited in space.

CCS graphic 1
A modularized RF front-end component; the package measures 6.5 mm x 5.4 mm.

The RF front-end begins at the antenna, reaches the RF transceiver and finishes at the modem. There's much more RF technology at play between the antenna and modem.

The diagram below attempts to simplify the many RF components in a leading 5G design. For suppliers of these parts, 5G poses a golden opportunity to expand their market as RF front-end content grows in proportion to the added RF complexities.

CCS graphic 2
A simplified schematic of the 5G RF front-end of the Samsung Galaxy S21; green boxes denote RF components and orange show millimeter wave antenna modules, which contain a transceiver, power management and phase array antenna.

The market for RF filters, for example, is poised to experience the highest growth rate within the 5G RF front-end, because support for more radio frequencies means more radio frequencies need to be filtered. Manufacturers of RF parts such as Broadcom, Qorvo, Skyworks Solutions, Murata and Qualcomm all stand to gain from this growth market.

However, supplying these RF front-end components in 5G phones is one thing, offering solutions to tame the RF complexity and make everything work is a whole different ball game. To address this emerging market need, Qualcomm, a traditional chipmaker, has put together a portfolio of RF components and technologies to offer a proven RF modem-to-antenna solution to smartphone-makers. This novel strategy brings significant value to the device ecosystem, because not all manufacturers want to invest large amounts of RF engineering resources to design their own RF front-end.

This article looked at how complex the RF front-end has become since the dawn of 4G LTE, exploring the driving forces behind that complexity. In the next two pieces in our three-part series, we'll uncover the key technologies at play in this area and highlight new opportunities for makers of RF components as 5G matures.

Wayne Lam is senior director, Research, Americas at CCS Insight. He has more than 12 years of experience covering mobile devices, wireless technology and associated supply chains. He joined CCS Insight after starting with smartphone pioneers such as Symbian and progressing to market intelligence firms including iSuppli.  

"Industry Voices" are opinion columns written by outside contributors—often industry experts or analysts—who are invited to the conversation by Fierce staff. They do not represent the opinions of Fierce.