3GPP Scout / Updates / News versus proposal

News versus proposalRAN1 #126

1Finity's AI interview and BG3 comparison have different scopes

Published 2026-10-16Updated 2026-10-163GPP TSG RAN WG1 #126

At a glance

1Finity's BG3 comparison appears in revised and not-treated contributions at RAN1 #126, with no agreement recorded.

What was announced and what was proposed

1Finity's Rob Hughes told RCR Wireless News: "MNOs can combine the AI services with the connectivity, the mobility, spectrum, transport, distributed edge". MNOs means mobile network operators, and AI means artificial intelligence. Our reading: the interview excerpt concerns combining services and connectivity, not the submitted base-graph comparison.

Source: RCR Wireless News, 2026-10-05, https://rcrwireless.com/20261005/ai/1finity-ai-ran-interview

RAN1 is the radio layer 1 working group, and both contributions identify RAN1 #126. Base graph 3 (BG3) labels designs compared by 1Finity, not a name unique to its proposal. 1Finity wrote: "Regarding row-orthogonality, compared with BG3 proposed by Huawei (RAN1-125 Huawei) and Samsung (RAN1-125 Samsung), our proposed BG3 (RAN1-126 1Finity) is much better." R1-2606098 contains the same observation, with identical wording. Row-orthogonality is the stated comparison, but the observation supplies no row counts or comparison method.

Source: R1-2606769, 1Finity, 3GPP TSG RAN WG1 #126. © 1Finity.

The register marks R1-2606769 not treated at RAN1 #126, with no decision recorded. Our reading: not treated means the meeting did not handle that document. The register marks R1-2606098 revised at RAN1 #126. Our reading: neither status records agreement on the BG3 comparison.

Sources

  1. R1-2606769, 1Finity, 3GPP TSG RAN WG1 #126. © 1Finity.
  2. R1-2606098, 1Finity, 3GPP TSG RAN WG1 #126. © 1Finity.
  3. RCR Wireless News, 2026-10-05, https://rcrwireless.com/20261005/ai/1finity-ai-ran-interview

How this page was made

Written with AI assistance from a news article and TDocs indexed by 3GPP Scout. An AI model (GPT-6 Sol) drafted each sentence and named the source it rests on. Code checked every quote word for word against the stored source text, and the live check fetched each TDoc again before publishing. A second AI model from another company (Claude Opus 5.5) reviewed the page against the same sources. The 3GPP documents themselves govern.