SoftBank and Ericsson have teamed up to improve 5G network performance by introducing advanced uplink switching technologies.
Historically, most commercial cellular networks were designed with a strong downlink bias. Operators allocated spectrum and configured time-division structures on the assumption that users consume far more data than they produce.
The rise of demanding mobile applications and local AI processing challenges that assumption. Modern usage patterns show devices continuously generating rich media and sensor data that must be uploaded to cloud processing with minimal delay.
To meet these new demands, SoftBank and Ericsson are deploying “Uplink Tx Switching” within standalone 5G networks. By optimising carrier aggregation together with multiple-input multiple-output (MIMO) techniques, this solution improves device-to-tower uploads and delivers more stable, higher-capacity connections.
Research highlighted in the November 2025 Ericsson Mobility Report identifies AI agents and AI-enabled devices as major drivers of increased upload requirements. As agents transmit ambient data, voice inputs, and high-resolution media for cloud processing, operators face a steep rise in uplink traffic.
Re-engineering resource allocation
Raising transmission capacity means using spectrum resources more efficiently. In prior configurations, SoftBank increased TDD uplink performance by aggregating carriers that included FDD bands, combining different frequency ranges to widen the overall data path available to devices.
Under the 3GPP Release 16 specifications, the approach has evolved toward dynamic radio resource reallocation. Uplink Tx Switching temporarily suspends the FDD connection at the exact moment a device transmits on a TDD carrier. Pausing the FDD link lets the smartphone’s RF frontend focus its transmit power and antenna resources exclusively on the TDD band.
This rapid reallocation enables devices to perform MIMO transmission across the broader bandwidth offered by TDD spectrum.
SoftBank validated this configuration in internal tests, including simultaneous transmissions across the n1 and n77 bands. Those trials indicate throughput gains approaching roughly 1.5 times current capacities.
Deploying such advanced physical-layer features requires close coordination throughout the telecom supply chain, because the user device’s hardware significantly affects attainable connection quality.
Acknowledging this dependency, SoftBank and Ericsson began working with semiconductor vendors in 2024. They conducted joint performance verification from the early stages of baseband chipset development to ensure the necessary RF switching logic is integrated into upcoming modems.
This extended lead time helps ensure smooth integration of the capability into SoftBank’s live network. Compatible smartphones are expected to start appearing commercially in summer 2026.
Enterprise economics and architectural value
Improved uplink performance enhances service reliability and creates new monetisation opportunities. Enterprises that manage remote crews, media production teams, or logistics fleets depend on consistent upload speeds. A carrier that can guarantee transmission of large data payloads can offer premium service tiers and influence average revenue per user (ARPU).
AI-driven workflows that offload local sensor or camera data to cloud platforms require sustained uplink capacity. If the network cannot support required egress rates, latency increases and applications fail to meet service expectations.
By optimising uplink architecture, operators can protect high-value enterprise contracts from degradation and support demanding industrial use cases, smart city systems, and computer-vision analytics without needing full physical densification of the radio grid.
Monetising enhanced uplink capabilities also requires updates to business support systems (BSS) and operational support systems (OSS). Telecom operators offering dedicated uplink slices to enterprises must ensure billing systems can accurately meter outbound traffic generated by specific AI processes.
Routing large uplink flows into multi-cloud environments calls for tight integration with hyperscaler edge zones so offloaded data reaches processing servers with minimal latency.
Tower companies are monitoring these radio access developments as well. As devices transmit heavier payloads more frequently, site-level power and cooling demands may fluctuate. Managing these factors is essential to preserve passive infrastructure margins.
Wholesale carriers may package access to networks with Uplink Tx Switching as a differentiated product, enabling bespoke IoT connectivity options for high-definition video surveillance or continuous telemetry collection.
Advancing towards Release 17
The current efforts based on Release 16 represent a milestone in a longer engineering roadmap. Standards bodies continue refining interaction models to extract greater efficiency from existing spectrum holdings. SoftBank plans to expand the device ecosystem and geographic coverage for this switching technology.
SoftBank and Ericsson will keep collaborating to drive further gains. Engineering teams are exploring ways to support MIMO across FDD bands, a capability outlined in 3GPP Release 17.
Delivering that functionality will require operators to balance legacy network investments while upgrading base station hardware to handle more complex baseband processing for diverse MIMO use cases.
Upgrading radio environments involves careful capital planning, negotiations with equipment vendors, and extensive integration testing to avoid disrupting existing services.
See also: Amazon Leo’s Globalstar deal changes the satellite race
Interested in learning more about IoT from industry experts? Consider the IoT Tech Expo, hosted in Amsterdam, California, and London. The event is part of TechEx and runs alongside other technology conferences such as AI & Big Data Expo and the Cyber Security Expo.
Telecoms content is produced by TechForge Media. Explore other enterprise technology events and webinars organised by the publisher.