Skip to main content
Satellites, NTN & HAPS · 3GPP Quarterly September 2025

Satellite IoT devices get a way to share channels, but collision rules still split vendors

Orthogonal cover codes now multiplex up to 4 devices per channel; Nokia and Apple want dropped codewords at gap boundaries, Ericsson and Huawei say scheduling can avoid it.

25 decisions17 work areas1,056 contributions

Plenary cycle

SA#109 · RAN#109 · CT#109

Beijing / China · 15 Sept 2025 – 19 Sept 2025

Source meetings for this article (9)

Meetings in this article’s source material. Inclusion does not mean that every meeting approved every decision.

Satellite IoT capacity now hinges on a fight over what happens at the edges of a transmission. Huawei, OPPO and Qualcomm want the scheduling pattern for multi-transport-block traffic scaled to match the orthogonal cover code length so devices sharing a channel don't step on each other; Nokia and Apple instead insist that any codeword touching a timing pre-compensation gap should be dropped outright, while vivo and Qualcomm counter that smart scheduling alone avoids the collision. Unlike terrestrial 5G, where interference is a capacity nuisance, on a satellite link with heavy Doppler shift it can silently erase the gains the whole feature was built to deliver.

The core win this quarter was confirming inter-slot orthogonal cover codes for satellite IoT: a code length of 2 multiplexes two devices, length 4 uses Hadamard sequences to multiplex four, and networks group devices by carrier frequency offset windows as tight as 50 Hz to keep the codes aligned. RAN1 also fixed a concrete edge case — when an uplink transmission collides with a random-access occasion, devices restart at a slot where (5×nf + ns) mod 4 = 0, keeping the 4-slot repetition pattern intact. Chipmakers get a stable rule instead of ambiguity, but three other pieces — dropping entire OCC groups on cancellation, CSI-report timing, and the beta-offset overhead cut — stayed unresolved despite broad support from Vivo, Samsung, DoCoMo and Panasonic.

A parallel study opened on what happens when satellite phones lose GPS entirely. 3GPP agreed to test two scenarios — total GNSS loss and gradual position drift — across geostationary orbit, 1200 km and 600 km low-earth orbit, at elevation angles down to 12.5 degrees and speeds up to 1500 km/h for aircraft terminals. But Apple, ST Engineering iDirect, InterDigital and CMCC want both cell-wide and per-device GNSS failure studied, while Nokia, Samsung, CEWiT and Lenovo argue only the per-device case matters; the split determines whether future fixes need network-wide signaling or just individual device logic.

NB-IoT over satellite got its TDD frame structure locked: 8 downlink subframes, a 50-subframe guard, 8 uplink subframes, then a 24-subframe guard, all within a 90-millisecond cycle. Nokia and Xiaomi pushed to shorten the scheduling delays to reclaim wasted resource inside that tight 8-subframe window, but Lenovo, Ericsson, ZTE, Huawei and Iridium — ten companies against two — kept the existing terrestrial delays, judging the change unnecessary in a maintenance release. A separate push by Qualcomm, Nordic and Iridium to relax reference-signal assumptions for better channel estimation was blocked by eight opposing companies including OPPO, Huawei and Nokia.

One dependency now blocks further progress: OPPO and Xiaomi want RAN1 to specify how devices pre-compensate timing across each block of 8 TDD uplink subframes, but Lenovo, Vivo, Huawei, Nokia and seven others say this belongs to RAN4, the group handling radio performance requirements. Until RAN4 weighs in, TDD-mode satellite IoT devices lack a finished timing procedure — a gap that matters for whether early chipsets ship with FDD-only assumptions baked in.

The decisions behind this

Deferred
Introduction of IoT-NTN TDD mode
R1-2506398

No new scheduling delays introduced for NB-IoT NTN TDD

Scheduling delays (k0) determine how many subframes a device waits after receiving a downlink grant before it can receive the data, or after an uplink grant before it must transmit. In the NTN TDD frame structure, with only 8 usable subframes in each 90ms period, some companies argued the standard delays are too large, wasting precious radio resources.

A proposal from Nokia to introduce a new set of shorter, more granular scheduling delays specifically for the TDD mode was discussed. Most companies, including Lenovo, Ericsson, ZTE, Huawei, and Iridium, opposed it, stating that the existing delays 'can work' and that introducing new functional changes is not necessary in the current maintenance phase of the standard. The feature lead noted that the issue was discussed earlier in the work item with no consensus to introduce changes.

For introducing new, shorter scheduling delaysAgainst new delays, keep existing ones
Nokia · Xiaomi2 companiesLenovo · Ericsson · OPPO · Vivo · ZTE · Spreadtrum · LG · CATT · Cambridge Consultants · Iridium · Huawei11 companies
Why it matters

Networks using NB-IoT NTN TDD must work with the existing, coarser set of scheduling delays defined for terrestrial NB-IoT. This may limit scheduling flexibility for satellite operators but maintains consistency and avoids new implementation complexity for devices.

Deferred
Non-Terrestrial Networks (NTN) for Internet of Things (IoT) Phase 3 (for LTE)
R1-2506591

Proposal to leave ambiguous SR and PUSCH overlap behavior to device implementation

A key detailed issue within the OCC framework involves handling conflicts when a device needs to transmit a Scheduling Request (SR) on a high-priority control channel while simultaneously having data scheduled on a shared channel (PUSCH) that is part of an OCC group. If the SR and PUSCH transmissions occur in different slots within the same OCC group, the rules for whether to drop the entire PUSCH group or multiplex the control information become ambiguous.

A group of companies, led by Ericsson and supported by Vivo, LG, OPPO, and others, proposed that in such a specific conflict scenario, the device's behavior should be left up to the manufacturer's implementation. They argue that the network can avoid the situation by ensuring the SR periodicity is not shorter than the OCC length. Other companies, including Samsung and Huawei, opposed this, stating that existing rules are sufficient and the PUSCH should simply be dropped, requiring no new specification text. The meeting moderator noted the discussion but no agreement was reached. The reason for the decision is not indicated in the meeting documents.

The moderator also put forward for discussion two related proposals from Spreadtrum. These aimed to restrict scenarios where a device would need to send multiple control channels within an OCC group, but these were also part of the unresolved debate.

Leave behavior to UE implementationUse existing rules, no spec change needed
Ericsson · Vivo · LG · OPPO · CATT · Panasonic · Spreadtrum · Apple (open to discuss)8 companiesSamsung · Huawei2 companies
Why it matters

If left unresolved, this ambiguity could lead to different behaviors between phone brands when connected to NTN. Networks would need to be designed cautiously to avoid the triggering scenario, potentially limiting scheduling flexibility. A future decision is needed to ensure consistent network performance and device interoperability.

  • DeferredDebate continues on whether to optimize reference signal availability for TDDR1-2506398
  • DeferredNo extra conditions for random access message repetitionR1-2506444
  • DeferredClarification proposed for time pre-compensation in TDD, but responsibility may lie with another groupR1-2506398
  • DeferredNTN IoT feature CB-Msg3-EDT moves forward with physical layer clarificationsR1-2506617
  • DeferredProposal on the nature of GNSS unavailability was not adopted, remains for further studyR1-2506614
  • DeferredCompanies propose but do not agree on DCI signaling changes for OCCR1-2506588
Quotations are verbatim from moderator summaries in the 3GPP archive. Document numbers link to the original files. Back to the full report.