- What’s new: Nokia and Rohde & Schwarz showcased an AI‑powered 6G receiver prototype at the Brooklyn 6G Summit this week. The proof‑of‑concept targets one of 6G’s toughest problems: coverage at higher frequencies. [1]
- Why it matters: Nokia says machine‑learning techniques can extend uplink distance by ~10–25% versus today’s receivers, enabling operators to roll out 6G over current 5G footprints, lowering cost and speeding time‑to‑market. [2]
- Where it runs: A research radio operating between ~7 and 30 GHz—aligned with the upper mid‑band/FR3 focus for early 6G—was shown on the expo floor. [3]
- Context: The demo caps a season of AI‑centric wireless news that includes NVIDIA’s $1 billion strategic investment in Nokia to accelerate AI‑RAN and the transition from 5G‑Advanced to 6G. [4]
What Nokia showed in Brooklyn
On the final day of the Brooklyn 6G Summit (Nov. 5–7, 2025)—co‑hosted by Nokia and NYU WIRELESS—Nokia highlighted HybridDeepRx, an AI‑enabled receiver designed to improve 6G coverage without a wholesale rebuild of cell sites. According to Nokia Bell Labs researcher Dani Korpi, the lab radio under test spans roughly 7–30 GHz, and pairs with an AI receiver to tolerate higher, distortion‑prone transmit powers—translating into a few extra dB of link budget. [5]
Nokia says the approach can boost uplink distance by 10–25% under realistic test conditions, a gain validated with Rohde & Schwarz gear (including an SMW200A vector signal generator and FSW signal & spectrum analyzer used for the AI inference workflow). Those improvements, if carried into products, could allow 6G to ride on existing 5G sites, cutting coverage‑driven capex. [6]
The HybridDeepRx demo is listed among the event’s featured exhibits, underscoring its role as a near‑term building block in AI‑native radio research. [7]
Why this matters: beating physics (with ML)
Higher‑frequency spectrum central to early 6G plans offers wider channels but shorter reach, forcing denser networks. Nokia’s AI receiver tackles the non‑linear distortion that creeps in when user devices push power amplifiers hard—typically a recipe for degraded error‑vector magnitude (EVM) and shorter uplink. By learning and compensating for those distortions (including digital post‑distortion techniques) the receiver recovers clean bits at longer ranges. That is the core of Nokia’s coverage claim and the premise for lower site counts in 6G overlays. [8]
Partnership details: Nokia × Rohde & Schwarz
Per the companies’ announcements, Nokia Bell Labs developed the receiver and ran validation using Rohde & Schwarz 6G test methods and instruments. The parties unveiled the proof‑of‑concept at B6GS on Nov. 6, positioning it as pre‑standardization technology aimed at accelerating 6G time‑to‑market. [9]
Light Reading’s Europe brief similarly flagged the tie‑up, noting Nokia’s claim that ML extends uplink distance by up to 25%, directly addressing FR3 coverage limits. [10]
A season of AI‑RAN momentum around Nokia
Beyond this week’s lab demo, Nokia has become a focal point for AI‑RAN investment. Late last month, NVIDIA and Nokia announced a strategic partnership to bring NVIDIA‑powered AI‑RAN products into Nokia’s RAN portfolio—paired with a $1 billion equity investment (about 2.9% stake) at $6.01 per share. The deal underlines the industry push to make AI a native part of the wireless stack from Layer‑1 up, and to bridge today’s 5G‑Advanced with tomorrow’s 6G. [11]
What’s next
- From demo to drafts: The receiver is still pre‑standard, but proofs like HybridDeepRx help shape 3GPP 6G requirements and trials expected to ramp later in the decade. Nokia’s Korpi also reminded that commercial 6G isn’t expected before ~2030—this is groundwork for that timeline. [12]
- Ecosystem validation: Expect more vendors and operators to reproduce the AI‑receiver results under multi‑vendor testbeds to understand coverage uplift vs. device power and network density trade‑offs.
- Cost calculus: If operators can reuse existing 5G site grids for wide‑area 6G, the AI receiver could become a high‑leverage feature for early deployments in FR3 bands. [13]
Key quotes (edited for brevity)
- Nokia Bell Labs on coverage: AI helps address “coverage limitations inherent in 6G’s higher‑frequency spectrum” by boosting uplink distance, with tests showing 10–25% gains. [14]
- Fierce Network from the show floor: Nokia’s AI receiver lets radios “detect higher transmit powers… which are distorted,” delivering a few dB of improvement in practice. [15]
Source notes & event context
- Fierce Network reported live from the Summit, detailing the 7–30 GHz test radio and the AI‑receiver concept. [16]
- Telecom Review Asia published specs and validation details, including SMW200A/FSW equipment and the 10–25% uplink‑distance claim. [17]
- Nokia and Rohde & Schwarz issued press materials confirming the Nov. 5–7 demo window and pre‑standardization status. [18]
- Light Reading (Eurobites) summarized the collaboration’s coverage goal and the up‑to‑25% distance improvement. [19]
- Reuters covered NVIDIA’s $1 billion Nokia investment for AI‑RAN and 6G progression. [20]
SEO snippets & FAQs
What is HybridDeepRx?
A Nokia Bell Labs machine‑learning receiver that compensates for non‑linear distortion, enabling higher effective uplink range at FR3‑like frequencies—demonstrated publicly at the Brooklyn 6G Summit 2025. [21]
How big is the coverage boost?
Nokia cites ~10–25% uplink‑distance improvement relative to conventional receivers in realistic tests using Rohde & Schwarz equipment. Real‑world performance will depend on spectrum, device power, environment, and network configuration. [22]
When does 6G arrive?
Commercial 6G is not expected before ~2030, but pre‑standard demos like this inform specs and early trials. [23]
References
1. www.nokia.com, 2. www.telecomreviewasia.com, 3. www.fierce-network.com, 4. www.reuters.com, 5. www.nokia.com, 6. www.telecomreviewasia.com, 7. www.b6gs.com, 8. www.rohde-schwarz.com, 9. www.nokia.com, 10. www.lightreading.com, 11. www.reuters.com, 12. www.fierce-network.com, 13. www.telecomreviewasia.com, 14. www.telecomreviewasia.com, 15. www.fierce-network.com, 16. www.fierce-network.com, 17. www.telecomreviewasia.com, 18. www.nokia.com, 19. www.lightreading.com, 20. www.reuters.com, 21. www.b6gs.com, 22. www.telecomreviewasia.com, 23. www.fierce-network.com
