5G phones will get MIMO channel data faster, using the simplest signaling path
Early CSI and SRS reports move through fixed MSG4 triggers, not periodic signaling; a separate bid to let antenna port densities differ within one array was rejected by 20 companies.
Source meetings for this article (9)
Meetings in this article’s source material. Inclusion does not mean that every meeting approved every decision.
- SA#10916 Sept 2025 – 19 Sept 2025 · Beijing · 3GPP ↗
- RAN#10915 Sept 2025 – 18 Sept 2025 · Beijing · 3GPP ↗
- RAN1#12225 Aug 2025 – 29 Aug 2025 · Bengaluru · 3GPP ↗
- RAN4#11625 Aug 2025 – 29 Aug 2025 · Bengaluru · 3GPP ↗
- SA3#12325 Aug 2025 – 29 Aug 2025 · Gothenburg Metropolitan Area · 3GPP ↗
- SA5#16225 Aug 2025 – 29 Aug 2025 · Gothenburg Metropolitan Area · 3GPP ↗
- RAN3#12925 Aug 2025 – 29 Aug 2025 · Bengaluru · 3GPP ↗
- SA2#17025 Aug 2025 – 29 Aug 2025 · Gothenburg Metropolitan Area · 3GPP ↗
- RAN2#13125 Aug 2025 – 29 Aug 2025 · Bengaluru · 3GPP ↗
Huawei, Samsung, Qualcomm, MediaTek and Nokia agreed on one thing this quarter and split hard on another. All backed triggering early channel measurements — antenna sounding and channel-quality reports — through a single, one-shot (aperiodic) signal right after a phone connects, rather than building in periodic or semi-persistent alternatives that Ericsson and Google wanted kept open for later study. But when vivo, InterDigital, Huawei, Spreadtrum and China Telecom proposed letting individual antenna resources within one aggregated array use different signal densities, they lost to a coalition of 20 companies including Samsung, ZTE, Qualcomm and Apple, who argued mismatched densities would corrupt precoding accuracy across ports.
The concrete win is a set of four approved low-density patterns for reference signals used in massive MIMO. Arrays aggregating 48, 64 or 128 antenna ports can now run at densities as sparse as 1/8 of a resource element per port, versus the denser legacy patterns — freeing spectrum that would otherwise carry pilot signals instead of user data. ZTE, Sony and Qualcomm flagged that some density combinations don't align cleanly with how older phones handle zero-power reference signals, a compatibility risk that shaped which combinations RAN1 actually approved.
Phones with three transmit antennas got their signaling finalized: the parameter 'fourPortSRS-3Tx' now officially tells a device to reuse a 4-port sounding signal but mute the fourth port, enabling 3T3R and 3T6R antenna-switching modes. This closes out multi-meeting work stretching back to RAN1#116 and gives chipset vendors a fixed target — though a companion question, how to pair phase-tracking reference signals with three antennas, split NTT Docomo, Samsung and Ericsson (favoring a simple unified rule) against ZTE, CATT and OPPO (wanting an explicit separate rule), and stayed unresolved.
Low-band carrier aggregation via switching — letting a single receiver hop between an FDD carrier and a downlink-only supplemental carrier — got its skeleton locked down: a 40-slot bitmap pattern per phone, switching gaps timed to hardware capability (35 or 140 microseconds), and a guarantee of at least one synchronization signal every 160 ms on each carrier. Ericsson objected that mandating that signal on the secondary carrier blocks the 'SSB-less' cell option introduced in Release 18, but the requirement passed anyway with ZTE, MediaTek and vivo's support.
The switching frequency itself is still unresolved: Spreadtrum, Samsung, Qualcomm and Xiaomi want hard caps like 8 switches per 40 slots, while Ericsson, Nokia and LG argue that limits this tight strip networks of scheduling freedom. That fight, plus an unsettled rule on whether phones or networks bear responsibility for avoiding synchronization-signal collisions during switching, carries into the next RAN1 session — the outcome will decide how much of this feature's real-world throughput gain operators can actually claim.
The decisions behind this
No agreements reached on several CSI enhancements for 5G Advanced in RAN1#122
The physical layer working group (RAN1) discussed several detailed technical enhancements for 5G Advanced's MIMO capabilities, specifically related to Channel State Information (CSI) reporting for up to 128 antenna ports. These discussions, documented in a moderator summary, did not result in any final agreements. Several proposals from companies were noted but not approved, meaning they will be revisited at future meetings. The key disagreements centered on how to precisely define the behavior of new CSI codebooks and reporting mechanisms for larger antenna arrays.
One major point of contention was a proposal (Proposal 1.B) to define how a 'soft scaling' feature interacts with an existing CSI feature (SD+PD NES). The proposal, led by Google, Fujitsu, Lenovo, and ZTE/Sanechips, was opposed by a large group including Samsung, Ericsson, Apple, Qualcomm, and Nokia. The moderator noted that the interaction between these features has not been explicitly agreed upon and requires more discussion. No reason for the opposition was recorded in the meeting documents.
Similarly, a proposal (Proposal 1.D) on antenna port mapping for a Doppler-based CSI codebook (Type-II) was presented by vivo, Apple, CATT, Xiaomi, and NEC. It faced opposition from another large group including Google, Samsung, Ericsson, and Qualcomm, who argued it would impact network implementation. The moderator deemed the proposal technically sound but noted the significant opposition. No reason for the opposition was recorded in the meeting documents.
A clarifying proposal (Proposal 1.F) regarding a bitmap used in CSI reporting for large arrays received support from Fraunhofer, Samsung, Ericsson, Apple, and others, but was opposed by NTT DOCOMO, Google, ETRI, ZTE/Sanechips, and Huawei/HiSi. The moderator called the proposal technically sound, but the meeting did not adopt it. No reason for the opposition was recorded in the meeting documents.
Another unresolved topic was a rule (Proposal 2.A) for associating interference measurement resources with channel measurement resources in a specific reporting mode. This proposal, supported by Huawei/HiSi, Google, and NTT DOCOMO, was opposed by Samsung, Ericsson, Apple, Qualcomm, and others. The moderator questioned the need for the rule. No reason for the opposition was recorded in the meeting documents.
A proposal (Proposal 1.G) to clarify the text for Sounding Reference Signal (SRS) port grouping was supported by Huawei/HiSi, NTT DOCOMO, vivo, Qualcomm, Samsung, and others, but was opposed by Google, Spreadtrum, and Nokia who saw no ambiguity in the current text. The moderator considered the proposal valid. No reason for the opposition was recorded in the meeting documents.
Conversely, one proposal (Proposal 3.A) for defining an equation for Coordinated Joint Transmission (CJT) calibration reporting received broad support from almost all listed companies, with only Nokia opposing it on the grounds that the standard is already clear. Despite this widespread support, the meeting did not adopt the proposal, and it was marked 'TO BE REVISED' for the next meeting.
| Supported the various specific technical proposals | Opposed the various specific technical proposals |
|---|---|
| Google · Fujitsu · Lenovo · ZTE/Sanechips · vivo · Apple · CATT · Xiaomi · NEC · Fraunhofer IIS/HHI · Samsung · Ericsson · Huawei/HiSi · NTT DOCOMO · Qualcomm · Spreadtrum · Fujitsu · ETRI · OPPO19 companies | Samsung · OPPO · NTT DOCOMO · Spreadtrum · vivo · ETRI · Ericsson · Apple · CATT · Xiaomi · Qualcomm · Nokia · Google · Samsung · OPPO · NTT DOCOMO · Spreadtrum · Fujitsu · ETRI · Ericsson · Lenovo · ZTE/Sanechips · NTT DOCOMO · Google · ETRI · ZTE/Sanechips · Huawei/HiSi · Samsung · OPPO · Spreadtrum · Fujitsu · vivo · ETRI · Ericsson · Lenovo · Apple · ZTE/Sanechips · CATT · Xiaomi · Qualcomm · Google · Spreadtrum · Nokia · Nokia44 companies |
The lack of agreements means the specification work for these 5G Advanced MIMO features is delayed. Mobile network equipment vendors and smartphone chipset designers cannot finalize their implementations until these technical details are settled. The recurring pattern of one or two companies blocking otherwise widely supported proposals suggests that resolving these issues may require higher-level arbitration in 3GPP.
Proposal adds equation for CQI calculation with linked delay reports
A coalition including Huawei, ZTE, and Google proposed adding a specific mathematical equation to the 5G standard to describe how a device should calculate Channel Quality Indicator (CQI) when a Coordinated Joint Transmission (CJT) report is linked with a separate delay calibration report. This scenario involves compensating for timing misalignment between different transmission points.
The 'Reason for change' stated is: 'It was agreed that CJTC Dd report can be linked with Rel-18 Type-II CJT... Analogous to Rel-19 Type-II CJT mode-1, this can be reflected in the CQI/PMI calculation equation. However, the current description in TS38.214 already captures the necessary behaviour albeit without equation.' Most companies supported the addition for clarity, though Nokia argued the specs were already clear without it. The moderator revised the proposal based on feedback from several companies.
The meeting did not make a decision on this proposal; it was noted and the text was revised for future discussion.
| Support adding equation for clarity | Believe current text is sufficient |
|---|---|
| Huawei · HiSi · ZTE · Sanechips · Google · Samsung · OPPO · Spreadtrum · Fujitsu · vivo · ETRI · Ericsson · Lenovo · Apple · CATT · Xiaomi · NEC · NTT DOCOMO · Qualcomm19 companies | Nokia1 company |
Adding the explicit equation may reduce implementation ambiguity for chipset vendors when combining CJT and delay reports, ensuring consistent CQI calculation across devices, which is critical for accurate link adaptation in advanced multi-point coordinated networks.
- DeferredEarly CSI acquisition for new connections and SCell activation will focus on aperiodic triggersR1-2506450
- DeferredLower CSI-RS density options proposed for high-port configurationsR1-2506450
- DecidedClarification on how a UE should interpret CSI-RS ports in the 5G-Advanced Doppler codebookR1-2506533
- DeferredClarification proposed for Type-I codebook basis selectionR1-2505535
- DeferredRule defined for interference measurement in CRI-based CSI for hybrid beamformingR1-2506533
- DeferredRAN1 makes progress on early downlink CSI acquisition for 5G MIMO Phase 6R1-2506576