GaN Power Amplifiers: Driving Sub‑6 GHz 5G Massive MIMO Beyond LDMOS

In the preceding entry of this series, we explored the Massive MIMO technologies propelling 5G rollout nationwide. While mmWave promises high‑bandwidth performance, the next few years will see 5G largely defined by Sub‑6 GHz deployments, demanding a leap in RF front‑end capability.

Engineers must deliver base stations that integrate RFFE more tightly, shrink in size, cut power use, boost output power, broaden bandwidth, sharpen linearity, and lift receiver sensitivity—all while meeting tighter coupling between transceiver, RFFE, and antenna. Only a compact, high‑efficiency, cost‑effective PA can satisfy these requirements and make Massive MIMO viable.

The evolution of RF power amplifiers has been dominated by LDMOS since the 1990s, especially below 2 GHz where its low cost made it the default choice. Gallium arsenide (GaAs) offered a competitive alternative at higher frequencies, but at the expense of lower power output and higher price. LDMOS secured the market during the 2G era, yet with 3G and 4G the efficiency plateaued. Despite advances such as Doherty topologies and envelope tracking, operators shifted to gallium nitride (GaN) during China’s 4G LTE rollout in 2014.

GaN, although newer, has become the preferred semiconductor for high‑power, high‑RF applications—exactly what Sub‑6 GHz 5G demands. Its superior output power, linearity, and efficiency have prompted OEMs to replace LDMOS with GaN. While LDMOS still dominates today’s RF base stations, GaN is poised to displace it in 5G Massive MIMO deployments.

GaN’s primary advantage is higher power density, a result of its wider bandgap that yields higher breakdown voltages. This enables more power per unit area, expanding coverage and allowing smaller, lower‑cost RFFE.

GaN power amplifiers can operate at temperatures up to 250 °F, beyond silicon’s limits. Their superior thermal conductivity simplifies heat‑sink design, reducing size and cost—critical for mobile network operators with hefty infrastructure budgets.

Higher efficiency cuts operational spend. In Doherty configurations, GaN reaches average efficiencies of up to 60 % at 100 W output, drastically reducing energy consumption for Massive MIMO arrays.

GaN also excels in wideband performance. While LDMOS tops 4 GHz, GaN can handle up to 100 GHz with five times higher power density. Lower parasitic capacitance and higher output impedance simplify wideband matching, enabling a single wideband PA to serve multiple Sub‑6 bands and reducing the need for separate narrowband radios.

Nonetheless, LDMOS remains a cost‑effective choice for certain frequencies, and GaAs still offers niche advantages. However, the industry’s shift to GaN underscores its strategic importance for Sub‑6 GHz Massive MIMO.

GaN’s adoption is accelerating across base stations and other sectors like defense and aerospace. Growing volume drives economies of scale, lowering price while delivering energy, size, and multi‑band benefits. Linearity is also improving; GaN is only in its second generation for base stations versus LDMOS’s fifteenth, and ongoing research promises even higher performance soon.

As barriers fade, system designers must master how to leverage GaN’s unique properties.

Embedded designers can tap GaN’s strengths, but must adopt material‑specific best practices. The next article in this series will unpack design guidelines, dispel myths, and chart GaN’s future beyond RF.

Related Contents:

• 5G and GaN: Understanding sub-6Ghz Massive MIMO infrastructure
• 5G’s biggest challenges for communications service providers
• How O-RAN will transform interoperability in 5G networks
• 10 key trends in wireless technology
• 5G roll-out: a marathon not a sprint

For more Embedded, subscribe to Embedded’s weekly email newsletter.