Preface:

At the industrial node where 5G evolves toward 5G-Advanced and 6G pre-research begins, patents have long surpassed mere intellectual property to become a core strategic tool for giants to carve out market territories, block competitors, and define technological roadmaps.

This war without smoke not only determines the survival of individual companies but also underpins the underlying logic propelling the entire RF front-end industry toward the next technological cycle.

Author | Fang Wensan

Image Source | Internet

Giants Use Patents to Define Territories and Lawsuits to Guard Boundaries

Last October, Skyworks and Qorvo announced a merger to create a U.S.-based high-performance RF, analog, and mixed-signal enterprise valued at approximately $22 billion, with the deal expected to close in early 2027.

This move warrants attention as it represents not just an ordinary acquisition but a concentration of patents and customer resources in the RF front-end industry.

Skyworks excels in mobile terminals, Wi-Fi, connectivity, and analog capabilities, while Qorvo leads in BAW, PA, filters, defense, connectivity, and power management.

Post-merger, the U.S. camp will boast a more complete patent portfolio covering power amplification, filtering, modules, packaging, and customer reach, intensifying pressure on Chinese RF front-end companies to upgrade to high-end markets.

On the global RF front-end patent landscape, the four giants—Skyworks, Qorvo, Broadcom, and Murata—have each staked their claims, continuously reinforcing their walls with new patents and repelling intruders through litigation.

Skyworks holds the largest patent portfolio, with approximately 9,800 patents covering PA design, envelope tracking technology, and module integration.

Its strategy leans toward "strategic alliances," collaborating with Qualcomm on PA envelope tracking and system calibration to share R&D costs while embedding its technical standards into Qualcomm's ecosystem.

This partnership has made Skyworks' PA solutions a near-default choice for Android flagship phones.

Qorvo pursues a "deepening moat" approach, with 6,100 patents centered on GaN high-power technology, wafer-level IC packaging, and temperature-compensated filters.

It extended GaN RF technology from military radar to 5G base stations, creating a formidable barrier with few short-term challengers.

In BAW filters, Qorvo holds just 12% market share but differentiates itself with LowDrift and NoDrift temperature-compensated SAW technology, establishing advantages in automotive and industrial sectors.

Broadcom's RF strategy is the most intriguing. Despite owning the thickest patent wall in BAW/FBAR, it is gradually exiting the pure RF market to focus on enterprise and infrastructure software.

This does not mean Broadcom abandons the value of RF patents; instead, its patent portfolio has become a "rent-generating asset."

After seven years of litigation, Nuos and Broadcom reached a cross-licensing agreement, demonstrating Broadcom's willingness to open patents under equitable conditions—but only for a fee and with strings attached, essentially monetizing its patent stockpile.

Murata adopts the most aggressive strategy. As the SAW filter leader, it rapidly expanded its patent portfolio, with a 5.57x increase in patent filings from 2017–2019, and shows no hesitation in litigation.

Its 2025 five-pronged lawsuit against Maxscend across China, South Korea, and Germany represents near-omnidirectional pressure.

The value of patent wars for giants lies here: defending profit pools in high-end markets while raising entry barriers for latecomers.

Patents may not directly generate revenue but extend product lifecycles, protect customer shares, and strengthen bargaining power in industry consolidation.

The Next Phase of Patent Battles Centers on Three Fronts

① Integration: L-PAMiD, a fully integrated module combining PA, LNA, filters, duplexers, and switches, is becoming standard in 5G smartphone RF front-ends.

In the era of discrete components, each part could be sourced and priced separately; in the module era, whoever provides fully integrated solutions holds pricing power and supply chain influence.

The patent barriers for L-PAMiD lie not in individual components but in integration design and packaging processes—areas where international giants have yet to establish absolute monopolies.

If domestic manufacturers secure patents during this integration window, they can transition from followers to rule-setters in the next competition.

② GaN (Gallium Nitride): GaN applications in RF front-ends currently focus on 5G base stations, satellite communications, and radar, with Qorvo, Macom, and Analog Devices leading in high-frequency GaN PA.

However, as GaN process costs decline and smartphone power demands rise, GaN PA's penetration into consumer devices is inevitable.

Domestic players like Sanan Integrated, Huatai Electronics, and Suzhou Dynax are advancing localized GaN RF solutions, presenting a rare opportunity window as the patent landscape remains fluid.

③ AI-Driven Intelligent RF: AI-enabled adaptive impedance matching, linearization calibration, and software-defined RF are shifting front-ends from fixed parameters to dynamic tunability.

End-to-end control latency dropping below 0.5 milliseconds allows real-time parameter adjustments during calls to adapt to environmental changes—unthinkable under traditional analog design paradigms.

Patent layout (strategic positioning) in this direction is just beginning. Whoever establishes an AI RF patent system first may define the product form (form) of RF front-ends in the next decade.

These three fronts share a common trait: they demand cross-component, cross-process, and cross-algorithm collaborative innovation, marking a turning point where patent wars shift from quantity races to quality competition.

Investment Perspective: Screening Domestic Players Amid Patent Barriers

The substitution of domestic RF front-end components has entered a new phase, shifting from volume expansion to direct competition in patent capabilities and system integration.

The patent war between Maxscend and Murata serves as a watershed. Instead of settling or circumventing, Maxscend launched invalidation proceedings and successfully overturned four of Murata's patent claims.

This proved Chinese firms' technical accumulation in filters has reached a level where they can confront giants head-on in patent nullification proceedings.

By late 2025, Maxscend had secured 170 patents (97 inventions), with 152 new filings in 2025 (109 inventions), primarily in high-end modules like RF filters.

It boasts one of the most complete patent systems in domestic RF front-ends, with its MAX-SAW filter performance nearing international standards and patents covering materials, structures, and processes.

With the Chipshow Semiconductor industrialization project, the company achieved vertical integration from design to manufacturing, creating a virtuous cycle between patent and process capabilities to deepen long-term barriers.

Fxiang Technology takes a different approach. By May 2025, it held 331 patents (162 inventions, 41 overseas), the highest overseas patent ratio among domestic RF firms.

While leading in domestic GaAs PA shipments, it needs to strengthen high-end modules and filters. Its investment logic: secure patents first, then fill product gaps.

Vanchip's story is more resilient. In 2025, it reported RMB 2.321 billion in revenue and RMB 44 million in net profit—its first full-year profit ever.

This turnaround stemmed from L-PAMiD modules entering flagship models and Phase 7LE Plus modules penetrating supply chains.

Its strategy: use products to drive patents, first commercialize modules, then build patent barriers through iterations.

The company focuses on GaAs processes, accumulating deep engineering experience and patents in 5G PA linearization and L-PAMiD integration. It is also expanding into WiFi 7 and satellite communications to unlock long-term growth.

Its 2025 RMB 270 million investment in Saibai Chuang perfected the filter industrial chain (supply chain), addressing weaknesses and gaining bargaining chips for future patent battles.

Onmicro represents another possibility. This Sci-Tech innovation board (STAR Market)-listed company has mass-produced L-PAMiD modules for domestic flagship phones and exclusive (exclusively) developed Tiantong satellite communication PAs.

Despite a RMB 65 million loss in 2024, its revenue surged, and product mix shifted from low-end discrete devices to high-end modules.

Its patents focus on L-PAMiD integration and satellite RF—areas where international giants lack absolute barriers.

Starry Semiconductor is the first domestic firm to release a fully self-developed LB L-PAMiD module with built-in TF-SAW duplexers, demonstrating the ability to integrate filters and PAs in a single module.

While Broadcom, Qualcomm, and Skyworks hold over 80% combined market share in L-PAMiD, Starry's breakthrough challenges the perception that China cannot develop fully self-developed L-PAMiD solutions.

Wisepower has carved out a differentiated path with its software-defined reconfigurable PA architecture, enabling single-chip multi-band adaptation via AgiPAM technology—a unique advantage in 5G-Advanced multi-band convergence.

This reconfigurable architecture, an original technical route, avoids many traditional PA patent traps while building its own patent moat.

The company holds over 600 PA-related patents (30% PCT international), leading in globalization layout (strategic positioning). Beyond smartphones, it has broken into Ericsson's 5G base station PA supply chain and completed Aerospace Fifth Academy (Aerospace Fifth Academy) verification for satellite PAs, expanding smoothly into B2B and high-end markets.

Kangxi Communications is a dark horse in WiFi FEM, with its core strength lying in the DeltaRF nonlinear PA architecture—a globally unique technical route.

This originality delivers performance advantages (30% efficiency gain, 25% power reduction) perfectly suited for WiFi 7's high-power demands while constructing a solid patent barrier.

As the only domestic RF firm in Qualcomm's WiFi 7 reference design, its performance is poised to accelerate with WiFi 7 adoption.

Conclusion:

This war, though smokeless, is exceedingly costly. It consumes R&D budgets, tests cash flows, prolongs product cycles, and weeds out players relying solely on component stacking and price cuts. From an industrial perspective, however, it marks a necessary stage for the RF front-end sector to mature.

Partial References: EE World: "RF Landscape Rewritten," JW Insights: "Defending Rights with Innovation, Competing Globally with Rules: Lessons from Maxscend vs. Murata Patent Litigation," EEWorld: "RF Front-End Module & Component IP Trends: Q1 2026 Monitor," Sina Finance: "Defending Rights with Innovation: Lessons from Maxscend vs. Murata Patent Litigation."