For harmonized 6GR design for TN and NTN, RAN1 studies to identify the technical aspects affected by NTN characteristics, as well as lessons learned from NR/IoT NTN

For the study of RAN1 6GR design, consider the minimum spectrum allocation in which 6G can operate, subject to further discussion and confirmation in RAN.

• Note: RAN4 involvement is necessary.

Identify the high-level aspects which impact on the NR-6GR MRSS support

• Including the lessons learned from LTE-NR DSS

On enhanced overall coverage, identify coverage target(s) considering diverse use cases and device types

Study and identify the lessons learned from NR duplex modes On 6GR duplexing study, RAN1 considers at least following duplex types

• FD-FDD
• Semi-static TDD
• gNB semi-static SBFD
• HD-FDD on UE side
• Dynamic TDD

Study whether to consider following duplexing types

• gNB dynamic SBFD
• UE SBFD
• gNB FD
• Note: Other duplex modes are not precluded

Study a scalable 6GR design for diverse device types, considering aspects:

• What should be commonly applicable to all 6G device types
• FFS: add-on features dedicated to specific device types, if any

Study and identify the lessons learned from NR spectrum utilization and aggregation framework

• DC is subject to RANP decision in June 2026
• Note: MRSS aspects are separate discussion

Study the device types from physical layer perspective to be supported by 6GR, subject to further discussion and confirmation in RAN

Study the following smallest maximum supported RF and BB UE BW without spectrum aggregation for at least one low-tier device type supported by 6GR framework from physical layer perspective, subject to further discussion and confirmation in RAN

• Opt1: 3MHz
• Opt2: 5MHz
• Opt3: 10MHz
• Opt4: 20MHz
• FFS: the UL bandwidth may be different to the DL bandwidth
• FFS: the bandwidth value may be different for different SCS, duplex modes, and bands.
• FFS: whether RF and BB UE BW are same or different

Identify the high-level aspects which impact on the 6GR sync signal structure and associated periodicity.

Study and identify the lessons learned from NR BWP framework

Note: High-level aspects to consider to enable lower CAPEX/OPEX with respect to current networks include, but not limited to

• UE/NW implementation complexity
• UE/NW energy efficiency
• MRSS
• Spectrum efficiency

RAN1 provides methodology and corresponding initial analysis of potentially achievable coverage to RAN#110 to determine the coverage target(s)

The aspects to consider for supporting NTN include, but not limited to

• Initial access, including cell search and SSB periodicity
• Coverage
• Duplexing
• Capacity
• Signalling overhead
• GNSS-less/resilient/based operation
• Large/varying doppler and propagation delay
• Beamforming / beam management / beam hopping

High-level aspects to consider for the 6GR sync signal structure include, but not limited to

• Sync raster design
• Spectrum allocation
• smallest maximum supported RF and BB UE BW without spectrum aggregation
• mobile broadband service requirements as high priority
• Energy efficiency for both BS and UE
• Detection/tracking performance, latency, and complexity
  • Including initial cell search
• Coverage target
• Common design for diverse device types
• Consideration of the supported deployment
• Consideration on whether the single sync signal structure is sufficient
• Note: Aspects impacting on the periodicity is to be discussed under AI11.5

If the minimum spectrum allocation is 3MHz with 15kHz SCS for 6GR,

• Opt1: Design of the common signals/channels (at least for SSB) for initial access by assuming bandwidth larger than 3MHz, which is applicable to any spectrum allocations with adjustment, if applicable
• Opt2: A single design of the common signals/channels (at least for SSB) for initial access by assuming minimum spectrum allocation as target bandwidth 3MHz, which is applicable to any spectrum allocations

For scalable 6GR design for diverse device types, RAN1 can at least consider the following, targeting applicable to all 6G device types,

• Basic initial access procedures from RAN1 perspective
• Other PHY features after initial access procedure, e.g., Other DL/UL control, scheduling/HARQ
• Coverage features to meet the identified coverage target
• Energy saving both at BS and UE sides
• MRSS
• Note: whether these features are supported, mandatory or optional is separate discussion

For the RAN1 study of “Re-use of existing 5G mid-band (~3.5GHz) site grid for 6G deployments in at least around 7 GHz and targeting comparable coverage to 5G mid-band”,

• The link budget template candidates 1 and 2 are used to calculate the metric(s) as starting point to compare existing 5G mid-band and 6G deployments in at least around 7 GHz, with potential future update.
• During initial access/random access
  • Coverage target is referring the bottleneck channel (i.e. Rel-15 NR Msg3) during initial access/random access for existing 5G mid-band
  • FFS target value(s) of data rate for data channels relative to 5G mid-band
• Following deployment scenarios are considered
  • Urban macro (both O2I and outdoor)
  • Sub-urban macro (both O2I and outdoor)
• Following carrier frequencies are considered to calculate the metric(s)
  • [4 GHz] as the existing 5G mid-band
  • 7 GHz as 6G deployment
• Template in R1-2509615 is to be used for collecting inputs on the values from companies.

For the smallest maximum supported RF and BB UE BW without spectrum aggregation for at least one low-tier device type supported by 6GR framework, from physical layer perspective, RAN1 to consider at least

• Overall device complexity
• Overall system performance impact
• Energy efficiency for both BS and UE
• Aim at a single common signals/channels design in idle mode and initial access for diverse device types, as well as meeting mobile broadband service requirements as high priority

Skeleton for TR 38.760-1 “Study on 6G Radio RAN1 aspects” v0.0.3 in R1-2509569 is endorsed.

Conclusion

• Template in R1-2506582 is to be used for collecting inputs from companies.
• Additional NTN or TN assumptions, if any, or any necessary change of the parameters, are to be incorporated into the updated one of R1-2506582.

• The deployment scenarios in TR38.914 should be considered for evaluation assumption
• The common evaluation assumptions including the antenna modelling, general system-level simulation assumptions (including the carrier frequency, bandwidth and subcarrier spacing used for link-level simulation) for the deployment scenarios in TR38.914, link budget and traffic models will be discussed in AI 11.2
  • Other assumptions including for link-level simulation specific to each technical topic will be separately discussed under each individual agenda.
  • Note: Subcarrier spacing decision is up to AI 11.3.2.

Study which of the following traffic models are to be used for 6G evaluations, e.g.,

• Full buffer
• FTP Model 1 (in TR 36.814)
• FTP Model 2 (in TR 36.814)
• FTP Model 3 (in TR 36.872)
• XR Traffic models (in TR 38.838)
• VoIP model (as in TR 36.814)

Study whether to introduce the following traffic models for 6G evaluations considering, e.g.,

• FTP-3 variant with packet delay budget requirement
  • Details FFS
• New traffic model considering a mixed/variable packet size and the associated time domain behaviors (e.g., time between adjacent packet arrivals, packet delay budget)
  • Details FFS
• New traffic model(s) considering the new use cases or services, e.g., AI/ML services, immersive communication services, etc.
  • Details FFS

For link budget template, consider the following candidates:

• Candidate 1: Reusing the link budget template from TR38.830, i.e., the following table with notes as follows:
  • The values of the parameters are TBD.
  • MCL in row (22bis) is TBD.
  • FFS: whether/how/why to update

• Candidate 2: Template as Table 7.10.1-1 from TR38.913.
  • FFS: whether/how/why to update.

Study traffic modelling for evaluations related to immersive communication services including but not limited to advanced XR [e.g., TR22.870] and haptics services,

• XR traffic models (in TR 38.838) are considered as starting point.
  • FFS the detailed modifications on the parameters to the XR traffic model, e.g., higher packet size, higher packet arrival rate, higher packet size deviation, PDB, etc.
• FFS how many models need to be defined and the corresponding representative use cases.
• FFS how to incorporate haptics traffic (TR26.854).

Send LS to SA4 requesting input if any on the relevant traffic characteristics, RAN1 can continue the study before SA4 potential response.

Study extensions to FTP Model 1/FTP Model 3 to incorporate the following:

• Multiple packet sizes and associated time-domain behaviors (e.g., inter arrival time)
  • FFS number of packet sizes (e.g., 2 or 3).
  • FFS whether to have fixed or variable packet size and packet arrival rate for a given UE.
  • FFS applicability of multiple packet sizes to only one or both of FTP Model 1/FTP Model 3.
  • FFS packet size and arrival rate characteristics.
• Packet delay budget (PDB) related parameters
  • FFS PDB applicability to packets (e.g., one PDB parameter for only one traffic flow or different PDB parameters for different traffic flows).
  • FFS how to consider the PDB, e.g., whether to drop packets when exceeding the budget, PDB aware metric.
• Note consider the following for PDB:
  • Applicability to the extension to FTP Model 1/ FTP Model 3 with one packet size.
  • Applicability or not to the extension to FTP Model 1/ FTP Model 3 with multiple packet sizes.

The attached templates for NTN in R1-2507956 are endorsed in principle.

Conclusion The following existing traffic models could be used for 6GR performance evaluations,

• Full buffer
• FTP Model 1 (in TR 36.814)
• FTP Model 3 (in TR 36.872)
• XR Traffic models (in TR 38.838)
• VoIP model (as in TR 36.814)
• Instant message (as in TR 38.840)
• Note that which model(s) will be used can be further decided when performing simulations in each individual topic.

For around 2GHz carrier frequency, for BS antenna modelling

Note1: A single TXRU is mapped per panel per subarray per polarization as mandatory option. Companies can provide results optionally, assuming fully connected TXRU mapping within a panel per polarization. Note2: Other combinations used in the simulation results are up to company to report.

The following configurations for system-level simulations could be used for 6GR evaluation:

Note: other simulation BW could be considered. Note: The layout for each scenario will be separately discussed, including the carrier frequency combination for single layer and/or two layers.

• For around 700MHz, 32 for total number of antenna element at base station, 4 for total number of TXRU at base station, (8, 2, 2, 1, 1; 1, 2) for (M,N,P,Mg,Ng; Mp, Np), and (0.5, 0.5)λ for (dH,dV) are assumed as the baseline combination.
• For around 700MHz, 64 for total number of antenna element at base station, 8 for total number of TXRU at base station, (8, 4, 2, 1, 1; x, y) for (M,N,P,Mg,Ng; Mp, Np), and (0.5, 0.5)λ for (dH,dV) are assumed as the optional combination. Note: Other values/combinations are up to company to report

For around 700MHz, for TXRU mapping at base station, it is adopted as mandatory option for simulation campaign that a single TXRU is mapped per panel per subarray per polarization. Note: Companies can provide results optionally, assuming fully connected TXRU mapping within a panel per polarization.

Final LS R1-2508184 is endorsed.

For around 4GHz carrier frequency:

Note1: A single TXRU is mapped per panel per subarray per polarization as mandatory option. Companies can provide results optionally, assuming fully connected TXRU mapping within a panel per polarization. Note2: Other combinations used in the simulation results are up to company to report.

For around 7GHz carrier frequency:

Note1: A single TXRU is mapped per panel per subarray per polarization as mandatory option. Companies can provide results optionally, assuming fully connected TXRU mapping within a panel per polarization. Note2: Other combinations used in the simulation results are up to company to report.

For around 30GHz carrier frequency:

Note1: A single TXRU is mapped per panel per polarization as mandatory option. Companies can provide results optionally, assuming a single TXRU is mapped per panel per subarray per polarization as mandatory option. Note2: Other combinations used in the simulation results are up to company to report.

At least the following carrier frequencies could be considered (from RAN1 perspective) for 6GR NTN evaluations:

• L-band (i.e., 1.5GHz)
• S-band (i.e. 2 GHz)
• Ku-band (FFS detailed frequency range)
• Ka-band (i.e. 30 GHz for UL, 20GHz for DL)

For the study traffic model(s) for 6GR AI/ML services:

• A representative AI/ML service is the generative AI, e.g., as defined in TR22.870. Send LS to SA4 (cc RAN2, SA1, SA2) requesting input if any on traffic characteristics for AI/ML services.

Note: RAN1 is discussing the following options for the model:

• Option-1a: The model is parameterized by Token, e.g., Token size, Token arrival rate, and Token delay budget.
  • Token is the minimum unit of data generated in the application layer.
  • How to associate Tokens to PHY layer packets.
  • How to reflect the variable importance of tokens.
  • Whether other parameters are additionally needed when tokens are encapsulated together into a packet, e.g., packet arrival rate, packet success rate, and packet delay.
• Option-1b: The model is characterized by the parameters of PHY layer packet, including e.g., packet size, arrival rates, latency requirement, reliability requirement, etc.
• Option-1c: reusing or extending the FTP-3/XR traffic model.
• FFS other models/options need to be defined for other AI/ML services.

Draft LS R1-2508183 is endorsed in principle.

For 6GR evaluation, the following are assumed for system-level simulation:

For 6GR evaluations related to immersive communications services, the following two amended XR models based on the existing XR traffic model (in TR 38.838) can be considered:

• Model-1: eXR model without Haptics
  • Regarding the statistical parameters for single stream CG traffic model defined in Table 5.4.1-1 TR 38.838, add values for immersive gaming regarding the data rate and the frame generation rate as in red:

- Regarding the statistical parameters for packet size following truncated Gaussian distribution in Table 5.1.1.1-1 TR 38.838, add values for immersive gaming regarding STD, Max, and Min values as in red:

- Regarding the statistical parameters for AR UL Model 1 defined in Table 5.5.2.1-1 TR 38.838, add values for UL-heavy video uploading regarding packet size, generate rate, data rate, and PDB values as in red:

- The jitter is modelled the same as XR traffic model.

• Model-2: eXR model with Haptics
  • Haptics traffic is defined as XR traffic packet generation with co-generated haptics packets.
    • FFS on how to generate the multi-channel haptics packet including how to handle silent periods of haptics and the haptics packet sizes.
    • FFS on how to co-generate haptics packets and the XR traffic packets.
  • Haptics packets has packet delay budget (PDB) of either 12 msec or 30 msec, which can be selected as a traffic model parameter.
• Send LS to SA4 to inform about the above agreement and check if SA4 has related inputs for the model.

Note: whether the working assumption can be confirmed relies on SA4’s response

Updating the BS antenna modelling agreed in the last meeting as follows:

• For around 700MHz carrier frequency, for BS antenna modelling,
  • update the (8, 4, 2, 1, 1; x, y) to be (8, 4, 2, 1, 1; 1, 4).
• For around 2GHz carrier frequency, for BS antenna modelling,
  • for outdoor combination 1 (i.e., 32AE/4TXRU), update the (M,N,P,Mg,Ng; Mp,Np) to be (8, 2, 2, 1, 1; 1, 2)
• For around 7GHz carrier frequency, for BS antenna modelling,
  • for outdoor combination 1 (i.e., 768AE/128TXRU), update the (M,N,P,Mg,Ng; Mp,Np) to be (24, 16, 2, 1, 1; 4, 16).
  • for outdoor combination 3 (i.e., 1536AE/256TXRU), update the (M,N,P,Mg,Ng; Mp,Np) to be (48, 16 ,2, 1, 1; 8, 16).

For 6GR evaluation, the layout for system-level simulation is assumed as follows:

• Note: Single layer will be prioritized for the evaluations.
• Note: The carrier frequency for the corresponding layout for the two layers will be reported by companies for the evaluations.
• FFS the minimum distance for random drop in two layers.
• Note: for system-level simulation of MIMO schemes, specific assumptions could be discussed under MIMO discussion

For 6GR evaluation, the total transmit power per BS for system-level simulation is assumed as follows:

Note: The values defined in option1 refer to the Report ITU-R M. [IMT-2030. EVAL]. The values defined in option2 is calculated based on the proportional scaling with simulation bandwidth under the limitation of the maximum BS Tx power of 56dBm.

For FTP Model 3, the packet delay budget (PDB) can be additionally considered,

• The latency characteristic of the traffic in RAN side (i.e., air interface) is modelled as packet delay budget (PDB). The PDB is a limited time budget for a packet to be transmitted over the air from a BS to a UE for DL, or from a UE to a BS for UL.
• For a given packet, the delay of the packet incurred in air interface is measured from the time that the packet arrives at the BS to the time that it is successfully transferred to the UE for DL, or from a UE to a BS for UL. If the delay is larger than a given PDB for the packet, the packet is said to violate PDB, otherwise the packet is said to be successfully delivered.
• Values for PDB, e.g., {10ms, 20ms, 30ms, 50ms, 100ms, 200ms, 300ms, 1000ms, 2000ms} can be considered.
• Which values will be used will consider the use case for the evaluations.

For 6GR evaluations related to Massive Communication (IoT),

• For comparability with 5G results and verify that 6G can meet the IMT-2030 connection density requirements, the mMTC traffic model from IMT-2020 (TR 37.910) may be used as a starting point. This traffic model can be applied in UL or DL.
• FFS: necessity of new traffic model(s) for 6GR evaluation in RAN1, e.g., for the following traffic types.
  • Triggered/polled reporting
  • Autonomous reporting (event-driven or periodic)
  • Remote actuation
  • Firmware/software upgrade

For 6GR NTN evaluations, the carrier frequency for Ku-band is 14GHz for UL and 11GHz for DL.

Final LS R1-2509596 is endorsed.

Draft LS R1-2509595 is endorsed in principle by adding TSG SA for CC.

For 6GR evaluation, the UE power class for system-level simulation is assumed as follows:

• FFS: 29dBm for 1 Tx..
• FFS: 31dBm, or 35 dBm or 43dBm with EIRP <55 dBm

For 6GR evaluation, the UE distribution and UE speed for system-level simulation is assumed as follows:

• UE number per TRxP will be dependent on the used traffic model. Other values can also be considered in the future evaluations.
• Other velocities and/or other outdoor/indoor ratio will be dependent on the used traffic model or use cases. Other values can also be considered in each of other topics.

For 6GR evaluations, RAN1 to consider BS antenna modelling for around 15GHz carrier frequency as follows:

Note1: A single TXRU is mapped per panel per subarray per polarization for Indoor combination 1 and Outdoor combination 1 and combination 2. A single TXRU is mapped per panel per polarization for Indoor combination 2 and outdoor combination 3. Note2: Other combinations used in the simulation results are up to company to report.

For traffic model(s) for AI/ML services, the following can be considered:

• Packet size:
  • How to model the packet size, a fixed one or multiple values, or modelled as a random variable.
• Packet arrival:
  • FFS the details to determine the packet arrival rate, e.g.,
    • N multiple packets arrive together as a burst. The burst interval time is modelled as a random variable.
      • Within the burst, the N packets arrive according to a statistical distribution.
    • Packets arrive separately.
  • FFS whether/how to model the Jitter and the relation with the packet arrival.
• FFS: Whether the packet importance is known. Whether/how to reflect the packet importance.
• Whether/How to consider the PDB, e.g., Packet delay budget: The latency characteristic of the traffic in RAN side (i.e., air interface) is modelled as packet delay budget (PDB). The PDB is a limited time budget for a packet to be transmitted over the air from a base station to a UE, or from a UE to a base station
• FFS Whether/how to consider the Packet success rate requirement: [xx%] and the relation with the PDB.
• FFS how to model different cases, e.g., image-based GenAI, video-based GenAI, and chatbot, etc.
• FFS: Whether/how other traffic models (e.g., XR, FTP1/3) can be used to reflect above characteristics.

Note: input from SA4 if any will be considered.

• For FTP3 extension with multiple packet sizes (the number of packet size X =FFS: 2 or 3), FTP 3-extension 1
  • For each packet size S_i, the packets arrive according to Poisson distribution (as FTP 3) with mean inter-arrival time T_i (or arrival rate λ_i where T_i = 1/ λ_i)
  • Y packet sizes are simulated for each UE
• Down-select one from following
• Alt1: Y=1; X=e.g., 2 or 3
• Alt2: Y=X; X=e.g., 2 or 3
• Alt3: Either Alt1 or Alt2 can be used depending on the evaluation purpose
  • FFS: values of S_i and T_i, and their inter-relation (if any)
  • FFS: change “packet size” to “File size” (terminology)
  • FFS timing relationship for different packet sizes if Y=X.
  • FFS the number of UEs for each of X different sizes in a drop if Y=1.
• Note: PDB can be considered separately if needed
• Note: modeling sessions with multiple packets in each session can be discussed separately if needed.
• Down-selection between X=2 and 3.

Regarding FTP3 extension with multiple packet sizes:

• The number of packet sizes \(X=2\);
• For each packet size \(S_i\), the packets arrive according to Poisson distribution (as FTP 3) with mean inter-arrival time \(T_i\) (or arrival rate \(\lambda_i\) where \(T_i=1/\lambda_i\));
• For Alt1: \(Y=1\) packet size is simulated for each UE
  • For FTP3-extension with \(X=2\), \(K\) is the ratio between arrival rates of the packet sizes, i.e., \(\lambda_1 = K \cdot \lambda_2\), with \(K>=1\), assuming \(S_1 < S_2\),
  • \(G>=1\) is the ratio between the number of UEs with packet size \(S_1\) and \(S_2\), respectively.
  • Values of \(S_i\), \(\lambda_i\), \(i=1, 2\), and \(G\) can be decided in evaluation phase.
  • Note: The following table is an illustration of the traffic configurations.

• For Alt2: \(Y=X=2\) packet sizes are simulated for each UE
  • The packet of each size is generated following the independent Poisson Process \((S_i, \lambda_i)\) with \(i=1,2\).
  • \(K\) is the ratio between arrival rates of the packet sizes, i.e., \(\lambda_1 = K \cdot \lambda_2\), with \(K>=1\) assuming \(S_1 < S_2\).
  • Values of \((S_i, \lambda_i)\) with \(i=1,2\) can be decided in evaluation phase.
  • Note: The following table is an illustration of the traffic configurations.

Define a 4 TXRU outdoor BS antenna configuration for about 4GHz carrier frequency as below.

For 6GR evaluation, RAN1 to model the radiation pattern of a single antenna element for FWA/CPE including,

• Candidate1: Isotropic,
• Candidate2: Directional with different half power beamwidth and maximum directional gains as described in Table 1 below,
  • CPE can be equipped with 1 to 3 antenna panels, each following Alt 1 \(((M,N,P, M_g,N_g; M_p,N_p), (d_H, d_v))\) configuration.
• Candidate3: Omnidirectional as described in Table 2 below.
• Note: The orientation of the CPE panel can be optimized.

Table 1: Directional radiation power pattern of a single antenna element for CPE

Table 2: Omnidirectional radiation power pattern of a single antenna element for CPE

For 6GR evaluation, RAN1 to model the UE antenna as follows for around 30GHz carrier frequency,

• UE antenna configuration follows Table 1 below.
• UE antenna radiation pattern follows Table 2 below.
• Other antenna configuration can be considered and up to companies to report.

Table 1: UE antenna configuration for around 30GHz

Table 2: UE antenna radiation pattern for around 30GHz [see Table A.2.1-8 TR38.802]

The agreed table for UE distribution and UE speed for system-level simulation, the further update is highlighted in cyan as follows:

~~FFS: Applicability for FWA~~ NOTE1: Regarding the number of UEs per TRXP, a smaller or the same number of UEs is assumed for each micro TRXPs compared to each macro TRXPs.

RAN1 to assume the UE antenna height and UE distribution for CPE for 6GR evaluations as follows: Note: Indoor and outdoor CPE pre-selection criterion or mechanism could be further discussed in the evaluation phase.

Conclusion: Evaluation assumptions for 6GR air interface (10.1)

The scenarios captured in TR38.914 but are not discussed in RAN1 are also candidate scenarios for RAN1 evaluations for 6GR.

Regarding the gNB transmission power assumptions in the evaluations, update the second note of the agreement as follows:

• Note: For evaluation purpose, BS Tx power scales up with bandwidth proportionally under the limitation of the maximum BS Tx power is 56dBm for outdoor and 33dBm for indoor for the above carrier frequencies.
• Note: The values defined in option1 refer to the Report ITU-R M. [IMT-2030. EVAL]. T~~he values defined in option2 is calculated based on the proportional scaling with simulation bandwidth under the limitation of the maximum BS Tx power of 56dBm.~~

For 6GR evaluation, RAN1 to model the UE antenna as follows for below 30GHz carrier frequency,

• Note: Each of other topics could further decide to use which combination(s) for the evaluations.
• Other combinations are not precluded for evaluations, e.g., 2T6R, 3T6R, 6T6R, 6T8R.
• Note: The antenna locations in Alt1 and Alt 2 in the following table are considered as examples and used for performance calibration.
• Any antenna array structures and/or antenna locations in section 7.3 in TR38.901 is possible for evaluations and up to companies to report.
• Note: The antenna locations in Alt 2 not included in section 7.3 in TR38.901 are up to companies to report.
• Note: The antenna element-wise power variation at the UE in TR 38.901 section 7.6.14.2 can be optionally considered for Alt2 for handheld devices.
• Note: The radiation power pattern of a single antenna element in Table 7.3-2 TR38.901 is assumed for Alt2.
• The isotropic radiation power pattern is assumed for Alt1 at least for handheld devices.
• Note: The antenna element/location of T is a subset of the element/locations for R.
• Note: The mapping between the combination and the device types might be separately discussed.

NOTE1: This combination is for IoT UE only. NOTE2: This combination is for CPE UE only. NOTE3: If number of TXRU and frequency combination is applicable.

For 6GR evaluations related to Massive Communication (IoT),

• In addition to the IMT-2030 L2 PDU message size of [32] bytes, evaluation of higher traffic loads can be used, e.g., evaluation with a larger L2 PDU message size of [320] bytes

Regarding the agreement on UE transmission power assumptions in system-level simulation, the further update is highlighted as follows in red:

• Transmission power of 35dBm is assumed for CPE only for below 30GHz.
  • Note: EIRP limit could be considered

NOTE1: It is up to company to report the simulated transmission power confined within the define peak EIRP range.

CP-OFDM waveform as defined in 5G NR is supported as the basis for 6GR for downlink

• Enhancements/modifications on CP-OFDM will be studied as potential additions
• DFT-s-OFDM or any other OFDM-based waveform will be studied as a potential additional waveform for downlink

Note: proponents to identify at least the target use cases, signals/channels to use the waveform, and how the proposal is intended (if applicable) to support multiplexing with CP-OFDM, including MRSS, and how multi-user multiplexing is supported, etc.

Note: Proponents are encouraged to provide more detailed information on their proposals for the next meeting, e.g.:

Proponents to characterize the main motivation for modification/additional waveform proposals:

• Targeted link direction, i.e. DL, UL or both
• Targeted use case (e.g. NTN, specific frequency range, etc.), if any
• Potential motivations metrics used, and quantified gains for a proposal, e.g.
  • Coverage
  • Network energy efficiency
  • UE energy efficiency
  • Spectral efficiency
  • High speed tolerance
  • Scheduling flexibility
  • Integration with ISAC

Proponents provide information on the following aspects, if applicable

• MRSS compatibility
• Target channels/signals, e.g. all channels, PxSCH only, etc.
• MIMO (SU and MU-MIMO) compatibility
• Target modulations, and impact to other modulations, if applicable
• Multi-user multiplexing/scheduling flexibility
• Multiplexing/coexistence with baseline waveforms
• Impact on synchronization and initial access
• Expected specification impact
• Transmitter/receiver complexity and impact to power consumption.

CP-OFDM and DFT-s-OFDM waveforms as defined in 5G NR are supported as the basis for 6GR for uplink

• Enhancements/modifications on CP-OFDM/DFT-s-OFDM will be studied as potential additions
• Other OFDM based waveforms are not precluded.

Draft LS R1-2508068 is endorsed with following revision:

1. removing “Additionally, if time permits, any feedback for CP-OFDM PAPR reduction/MPR values achievable by implementation is also appreciated.”

• For uplink low-PAPR proposals, the link level performance evaluation criterion is Net Gain assuming same spectrum efficiency as the reference
  • Net Gain [dB] = Tx power gain relative to the reference – SNR degradation relative to the reference @10% BLER
    • A realistic PA model should be used
    • When calculating the Tx power gain, the RAN4 metrics on the Tx power should be taken into account.
    • For SNR degradation, fading channel and non-ideal channel estimation, including DMRS configuration, and equalization is encouraged.
    • FFS: Other evaluation metrics
    • Note: Companies to report how to calculate the Tx power gain, modulation and coding

• Study the evaluation method for evaluating DFT-s-OFDM for UL with number of layers > 1.

Final LS R1-2508069 is endorsed.

Table: System level configuration for multi-layer UL CP-OFDM/DFT-s-OFDM study: 

Add the following metrics for UL PAPR reduction to the existing agreement (made in RAN1#122b)

• Net gain@10% BLER assuming similar spectral efficiency and same occupied bandwidth for each compared method
• ACLR, EVM, IBE

General evaluation assumptions for UL low-PAPR proposals

General evaluation assumptions for UL low-PAPR proposals

For single user evaluation assumption for MCS and subcarriers UL low-PAPR proposals with spectrum extension

For single user evaluation assumption for MCS and subcarriers UL low-PAPR proposals with spectrum truncation

Note: other values for extension or truncation are not precluded.

For both CP-OFDM and DFT-S-OFDM, for single user evaluation assumption for MCS and subcarriers UL low-PAPR proposals with tone reservation.

Table is endorsed to characterize each proposal as a potential RAN1 observation. Characterization of each waveform proposal 

For UL PAPR reduction, values for occupied BW B: - {2, 4, 8, 16, 24, 30, 32, 64, 128, 240, 256} PRBs. - Other PRB allocations are not precluded. - Edge, outer and inner PRB allocations as defined in TS 38.101 should be considered. ## Agreement: Waveform (11.3.1) - Performance benefit to be evaluated using both link level and system level simulation. - FFS: metrics - Link level configuration for multi-layer UL waveforms study. 

R1-2601512 Session Notes of AI 10.2.1 Ad-Hoc Chair (NTT DOCOMO, INC.) Session notes are endorsed and incorporated the session notes below.

Extend the RAN1#123 endorsed table to characterize each (waveform) proposal as a potential RAN1 observation as follows to cover also impacts to transmitter and receiver processing operation:

Conclusion: Waveform (10.2.1)

DFT-s-OFDM waveform including related enhancements for 6GR Downlink will be no further discussed as part of AI 10.2.1.

• Note: for DL signal (e.g., SS, WUS, sensing), it may or may not be separately discussed in corresponding AI.

Conclusion: Waveform (10.2.1)

Studies on UL coverage improvements through low UL PAPR enhancement for DFT-s-OFDM are to be handled with high priority in AI 10.2.1.

Conclusion: Waveform (10.2.1)

Studies on DFT-s-OFDM for multi-rank UL MIMO are to be handled with high priority in AI 10.2.1.

For the evaluations of spectrum extension and spectrum truncation for UL low-PAPR solutions, the number of subcarriers A before extension / truncation is

• Option 1: \(12 * 2^x * 3^y * 5^z\) subcarriers
• Option 2: \(2^x * 3^y * 5^z\) subcarriers
• FFS: whether the maximum value for A is needed Note: Companies are encouraged to investigate above options and bring inputs to RAN1#124bis. Note: the occupied bandwidth B is given in terms of number of RBs.

NR Rel-15 DFT-s-OFDM should be used as the baseline reference when evaluating the gains of UL low-PAPR proposals.

Following metrics are used for SLS evaluations for multi-layer UL DFT-s-OFDM and CP-OFDM studies.

• User perceived throughput (UPT), including:
  • High percentile (90%)
  • mean
  • median
  • cell edge (5 & 10-percentile)
• Optional for full buffer traffic only: cell average throughput
• Companies are encouraged to report the CDF of instantaneous UL TX power across all UEs
• Companies are encouraged to report the statistics on the UL TX rank.
• Companies are encouraged to report the statistics on the applied MCS.

Following metrics are used for LLS evaluation for multi-layer UL DFT-s-OFDM and CP-OFDM studies.

• BLER curves (for a subset of NR MCS (covering whole range with spanning), HARQ re-transmissions disabled) for same transmission rank, for same resource allocation and same transmission power for DFT-s-OFDM and CP-OFDM
  • Companies may derive Link-level throughput vs SNR based on BLER curves
• Netgain

For 2-layer/2Tx UL DFT-s-OFDM and CP-OFDM studies, following two cases are considered.

• UE capable of only non-coherent precoder
  • For UL CP-OFDM and DFT-s-OFDM rank-2 transmission, only non-coherent precoder option is allowed.
• UE capable of fully-coherent precoder
  • For UL CP-OFDM rank-2 transmission, all NR precoder options are allowed.
  • For UL DFT-s-OFDM rank-2 transmission, only non-coherent precoder option is allowed.

For UL DFT-s-OFDM rank-2 transmission, only non-coherent precoder option is allowed. For the multi-layer UL DFT-s-OFDM and CP-OFDM studies, the NR reference should be evaluated assuming the Release 16 full power mode 1 to be enabled.

The following UL low-PAPR proposals for DFT-s-OFDM are for further consideration:

• Non-AI-ML-based
  • FDSS (R1-2601092, R1-2600751, R1-2600801, R1-2600823, R1-2600914, R1-2601156)
  • FDSS - spectrum extension (R1-2600027, R1-2601092, R1-2600823, R1-2600751, R1-2600914, R1-2601156, R1-2600261)
  • FDSS - spectrum truncation for \(\pi/2\) BPSK (R1-2601268, R1-2601212, R1-2601092, R1-2601156)
  • Tone Reservation (R1-2600261, R1-2600716, R1-2601268)
    • For \(\pi/2\) BPSK or other modulation orders
  • GMSK-Approximation based FDSS (R1-2600823)
  • 3-tap filter based FDSS (R1-2508684)
  • CFR-SE (R1-2600499)
  • Offset-QAM or \(\pi/2\)-PAM with FDSS-spectrum truncation (R1-2600909, R1-2600138, R1-2601268, R1-2600751)
  • Offset-QAM or \(\pi/2\)-PAM with FDSS-spectrum extension (R1-2600909, R1-2600138, R1-2600751, R1-2600823)
  • Interpolation-based modulation (R1-2600261)
  • AFDM based on DFT-s-OFDM (R1-2600999, R1-2601019)
• AI-ML-based
  • AI/ML-based waveform (R1-2600499, R1-2600751)

Note: tdoc numbers described in each proposal provide information on the proposal for further consideration, and it does not mean these tdocs support the proposal.

Conclusion: Waveform (10.2.1)

Further clarifications on evaluations of UL low-PAPR proposals for DFT-s-OFDM:

• In the evaluation assumptions, companies should disclose the necessary knowledge (e.g., filter coefficients, extention scheme, ...) of the waveform at the receiver to process if any.

• 6GR takes the following SCS as start point for discussion for all the signals/channels except PRACH.
• For sub 6GHz
  • The following subcarrier spacing is at least supported
  • 15kHz SCS for FDD, 30kHz SCS for TDD
  • FFS: 30kHz SCS for FDD for around e.g., 1-2.5GHz
  • FFS: 7.5kHz SCS for sub1GHz (FDD)
  • Whether to discuss the FFS will be subject to RANP decision.
• For around 7GHz
  • The following subcarrier spacing options can be studied
  • 30kHz, 60kHz
• FFS: For around 15GHz
  • The following subcarrier spacing options can be studied
  • 30kHz, 60kHz, 120kHz
  • Whether to discuss it will be subject to RANP decision
• For between 24.25GHz - 52.6GHz
  • Subcarrier spacing 120kHz is supported
  • FFS whether to allow using additional subcarrier spacing for SSB
• FFS subcarrier spacing for PRACH and up to initial access discussion.

• 6GR supports normal cyclic prefix, i.e., same as the normal CP defined in NR.
  • FFS potential need for other CP

Conclusion

• Numerologies for sensing is up to sensing agenda discussion.

6GR study assumes same SCS between 6GR Sync signals and other channels/signals (except PRACH) for a given band.

• FFS: same/different SCS between 6GR sync signal and other channels/signals (except PRACH) for FR2- 1.
• Note: ISAC is separately discussed in ISAC session.

• RAN1 assumes 400MHz maximum channel bandwidth at network side and 30kHz SCS around 7GHz
  • Study whether and how to enable UE to support 400MHz bandwidth

For communication, 6GR considers NR frame structure used as a starting point for the study item,

• Resource defined by one subcarrier and one symbol is called as resource element (RE).
• Resource block (RB) is defined where the number of consecutive subcarriers per RB is the same for all numerologies and the number of subcarriers per RB is 12
• Radio Frame length is 10ms
• Each radio frame is split into 10 subframes, each with a duration of 1 ms
• For given SCS and for given symbol, the symbol duration, normal CP length and boundary is same as NR design.
• A slot is defined as supporting 14 consecutive symbols for normal CP case and all subcarrier spacings.

For how to enable UE to support 400MHz bandwidth when a network supports 400 MHz Channel Bandwidth (CBW), the following options 1/2/3/4/5 are considered from RAN1 understanding for studying

6G HOW NOTE: IMAGES ARE NOT SUPPORTED YET. CHECK THE NOTES.

• Option 5: Variance of Option 3 by assuming single FFT and 2 RF chain.
• FFS which aspects of the BB processor in option 3 and 4 should be separated/parallelled.
• Note: DL and UL design options may be considered independently.
• To provide potential specification impact of each option.
• To provide investigations on performance/energy efficiency/cost/complexity for the above options.
• Inform RAN4 about the above information.

• RAN1 assumes maximum channel bandwidth 800MHz or 400MHz at network side for FR2-1
  • 800MHz or 400MHz, to be down-selected in the future
• FFS: 800MHz or 400MHz at UE side.

Draft LS R1-2509577 is endorsed in principle

Final LS R1-2509578 is endorsed.

Chairman Guidance For 6GR data channel coding,

• Evaluations can be provided in form of BLER results.
• Evaluation/analysis on throughput, complexity, and decoding latency can be provided
  • Other metrics are not precluded.
• Proponent companies to provide their target scenarios and requirements, evaluation assumptions and methodologies for respective evaluation/analysis, e.g., decoding algorithm and details, information sizes, code rates, HARQ scheme, channel type, modulation order, target BLER, etc.
• Proponent companies to provide details of channel coding extension compared with NR channel coding.
• Proponent companies to provide justification for the channel coding extension, and how to satisfy 6G requirements and characteristics with acceptable performance/complexity trade-off, compared with data channel codes as defined in 5G NR.

Chairman Guidance For 6GR control channel coding,

• Evaluations can be provided in form of BLER and FAR results.
• Evaluations/analysis can be provided for complexity, decoding latency,
  • Other metrics are not precluded.
• Proponent companies to provide evaluation assumptions and methodologies for respective evaluation.
• Proponent companies to provide details of channel coding extension compared with NR channel coding
• Proponent companies to provide justification for the channel coding extension, compared with control channel codes as defined in 5G NR.

• Study 6G data channel coding for higher throughput than 5G with acceptable performance-complexity tradeoff for both NW side and UE side,
  • Target peak data rate is assumed to be 2 times of the target peak data rate defined in TR38.913 Note: The other target throughput is up to company to report. Note: Applicability of the potential channel code will be further discussed.

• For 6G channel coding, LDPC is used for data (including SIBs) and Polar code is used for L1 control information (larger than 11 bits, including PBCH)
• For 6G LDPC
  • Working assumption: For data rate within NR range, reuse of NR LDPC design is supported
  • For data rate beyond NR range, study LDPC extension with acceptable performance-complexity tradeoff for both NW side and UE side
    • Note: Applicability of the potential LDPC extension to data rate within NR range will be further discussed
• For 6G Polar code
  • Working assumption: For control information within NR range (larger than 11 bits), reuse of NR Polar code design is supported
  • For control information beyond NR range, study Polar code extension with acceptable performance-complexity tradeoff for both NW side and UE side
    • Note: Necessity for control information beyond NR range is to be further discussed
    • Polar code maximum mother code length is kept as 1024.
• FFS: further motivation(s) for potential extension/enhancement until RAN1#123

For Polar code design for UCI with payload size larger than NR range (i.e., larger than 1706 bits), at least the following option is identified for further study

• More than 2 segments Note: The necessity of UCI payload size larger than NR range needs to be confirmed by other agenda(s)

Conclusion: No consensus on motivation(s) for potential extension/enhancement for Polar code design with payload size within NR range (larger than 11bits).

For the study of BG(s) and PCM(s) for LDPC extension for data rate beyond NR range, at least the following evaluation assumptions will be considered.

• Computation complexity is defined as (the number of iteration times for required BLER) *(the number of ones in the lifted parity check matrix)/ (CB size)

• For the study of channel coding for small UCI with payload size of 3~11bits, at least considering:
  • 5G RM code
• Identify the justifiable drawbacks of 5G RM code, if exists, study potential solution(s).

For the study of LDPC extension for data rate beyond NR range with acceptable performance-complexity tradeoff,

• To provide the initial version of LDPC BG(s) and PCM(s) in the excel spreadsheet by RAN1#124
• To provide the required SNR and complexity for target BLER, and the evaluation assumptions of the decoding algorithm
  • The definition of complexity will be further discussed
• FFS: other metrics

For the study of LDPC extension beyond NR range, the following options of the number of information columns in BG are identified in RAN1#124 for further study

• Option 1: 22
• Option 2: 44
• Option 3: 33

For the base graph of LDPC extension beyond NR range, the following is identified in RAN1#124 for further study:

• The LDPC extension is quasi-cyclic LDPC codes, the parity check matrix of quasi-cyclic LDPC Codes is defined at least by a matrix \(H\) of size \((mb \times z) \times (nb \times z)\), which consists of sub-block matrices of size \(z \times z\), where each sub-block matrix is composed by circularly shifted matrices or zero matrices. Wherein, \(mb\), \(nb\) and \(z\) are integers larger than 1.
  • The values of \(mb\), \(nb\) and \(z\) are FFS.
• The row weight of circularly shifted matrices is less than or equal to 2.
  • FFS: Whether row weight of all circularly shifted matrices is 1, or the row weight of some circularly shifted matrices is 1 and the row weight of some circularly shifted matrices is 2.

For the study of LDPC extension beyond NR range, the following options of the maximum code block size are identified in RAN1#124 for further study

• Option 1: 8448
• Option 2: 8448 * 2

For the study of LDPC extension beyond NR range, the following options of puncturing before rate matching are identified in RAN1#124 for further study

• Option 1: no puncturing
• Option 2: puncture 1 column
• Option 3: puncture 2 columns

For study of LDPC extension beyond NR range, at least the following metrics are considered for the performance comparison among BG1 and \(BG(s)/PCM(s)\) proposed by companies

• Computational complexity difference to achieve the same BLER performance as BG1 under the reference maximum number of iterations of BG1.
• SNR performance difference with the same computation complexity as BG1 under the reference maximum number of iterations of BG1.
• The reference maximum number of iterations is 5, 7, 10, 15 based on BG1.
• Other values can be also reported by companies Note: Companies can report the average number of iterations difference to achieve the same performance as BG1 under the reference maximum number of iterations of BG1.

For performance evaluation purposes, the corresponding evaluation assumptions for study the Polar code segmentation enhancement used for L1 uplink control information with payload size beyond NR range and larger than 1706 bits, following evaluation assumptions are considered.

Note: 11 bits CRC per segment, rate matching, channel interleaver and code-block concatenation follow NR. Note: Companies to report other values that are used in their evaluations.

For the study of LDPC extension beyond NR range, the maximum lifting size is not larger than 384

• FFS: the exact value of the maximum lifting size

• For 6GR DL, 5G NR uniform QPSK, 16QAM, 64QAM, 256QAM and 1024QAM are supported as basis for study for data channel
  • FFS: Enhancements and other modulation schemes
• For 6GR UL, 5G NR uniform QPSK, 16QAM, 64QAM, and 256QAM are supported as basis for study for CP-OFDM for data channel
  • FFS: Enhancements and other modulation schemes
• For 6GR UL, 5G NR pi/2 BPSK, uniform QPSK, 16QAM, 64QAM, and 256QAM are supported as basis for study for DFT-s-OFDM for data channel
  • FFS: Enhancements and other modulation schemes

For 6GR constellation shaping evaluation for CP-OFDM, and improved MCS table, the proposed scheme will be compared with non-shaping with NR MCS table. The evaluation and comparison should consider at least the following:

• BLER performance under AWGN channel (at least for performance calibration)
  • 1st transmission (baseline) and with HARQ re-transmission
• BLER performance under fading channel with fixed MCS
  • 1st transmission (baseline) and with HARQ re-transmission
• Throughput performance with link adaptation (adaptive MCS and rank) under fading channel
  • Needs to provide assumptions on rate adaptation (e.g., target BLER for 1st transmission, maximum # of retransmissions)
• Transmitter and receiver complexity (e.g., shaping/deshaping, demapper), latency, parallelism implementation, and storage requirements,
• Other KPI not excluded, such as PAPR, EVM, MPR/A-MPR
• Expected spec impact
• FFS detailed assumption of constellation shaping and improved MCS table
• System level evaluation can be done after link level evaluation.

For the study of uniform 4096QAM for DL and uniform 1024QAM for UL, need to study performance (assuming realistic channel estimation, time/freq synchronization assumption, phase noise assumption, etc), complexity/power consumption, requirements, benefit/necessity under applicable scenarios, associated restrictions, and challenges (such as EVM requirement, PAPR increase, MPR or A-MPR increase under realistic PA model).

• FFS: How to involve RAN4 early
• FFS: Shaping of higher order modulation
• System level evaluation can be done after link level evaluation.

For link level simulation for modulation evaluation, companies are encouraged to evaluate with the following assumptions and should report the exact scheme evaluated.

• channel configuration, including Channel profiles,Tx/RX antenna settings
• For MIMO scenario: SU-MIMO or MU-MIMO, follow agenda item 11.2 for MIMO when available.
• Precoder assumption
  • Close loop MIMO (reciprocal beamforming (e.g., SVD, SLR/RZF, etc.), codebook based)
    • Realistic CSI/SRS/AP-SRS periodicity and delay, and SRS chanEst assumptions,
    • or genie beamforming
  • Open loop MIMO
• Receiver assumption (for MIMO): LMMSE (baseline) for UL, rML or LMMSE for DL
• LLR demapper: Max-log (baseline) or Log-MAP
• Channel estimation: Realistic (baseline) or ideal
• Other assumptions: Channel coding NR-LDPC (baseline), PxSCH bandwidth, SCS, FD interleaver used or not, 5GNR BICM interleaver usage
• Note: For MIMO, SIMO, MISO and SISO are included when possible

For 6GR constellation shaping study, proponent is encouraged to provide details for the PS/GS schemes considered for evaluation and comparison, including at least the following

• Probabilistic shaping for CP-OFDM and DFT-s-OFDM
  • Use the list of spectrum efficiencies in NR MCS table as starting point, and provide constellation (including normalization), coding rate and target probabilistic distribution for each SE
    • If multiple coding rate and target probabilistic distribution pairs are provided for each SE, how to switch between them
  • Relationship between shaping and FEC, coded bits to modulation symbol mapping, and other modules (such as scrambling, interleaving), in transmit and receive chains. How to handle HARQ retransmission
    • PS algorithm details (for example, source coding based, channel coding based, etc) and parameters (such as block length, rate loss)
• Geometric shaping for CP-OFDM and DFT-s-OFDM
  • Use the list of spectrum efficiencies in NR MCS table as starting point, and provide target constellation shape (including normalization) (1D-NUC, 2D-NUC, QAM-CS, etc) for each SE
    • If multiple constellation shapes are provided for each SE, how to switch between them
  • GS mapping details, such as bit to constellation point mapping and shape
  • Relationship with other blocks (such as scrambling, interleaving). How to handle HARQ retransmission

For 6GR constellation shaping evaluation for DFT-s-OFDM, and improved MCS table, the proposed scheme will be compared with non-shaping with NR MCS table. In addition to what has been agreed for CP-OFDM in earlier agreement, the evaluation and comparison should further consider at least the following:

• PAPR/CM of the resulting waveform
• EVM, MPR/A-MPR

On how to evaluate complexity, storage requirement, delay and parallelism/serialism for PS and GS compared with uniform QAM.

• For PS
  • The demapper complexity is compared with uniform QAM demapper complexity
    • Can report the demapper complexity with PS and the demapper complexity of NR MCS with the same spectrum efficiency, and the ratio of the complexities
      • Also report the number of spatial layers, dm-algorithm used and the receiver type (e.g., LMMSE or rML), and fixed point assumed or floating point assumed.
  • The Distribution Matcher (DM)/Distribution De-Matcher (DDM) complexity and/or storage requirement as a function of the DM algorithm used (ESS, CCDM, etc), precision of fixed point implementation, block length, and the number of bit levels shaped per symbol
    • For complexity, can report the complexity normalized by the number of information bits
      • As a reference, can also report the computation complexity of LDPC decoding with 10 iterations.
    • For storage requirement, can report the overall storage needed for DM/DDM for supporting all MCS in the MCS table and all shaping related parameters
    • DM and DDM complexity and storage will be counted separately
  • The DM/DDM processing delay, parallelism/serialism, as a function of DM design and block length, and their impact to throughput
• For GS,
  • The demapper complexity is compared with uniform QAM demapper complexity
    • Can report the ratio of GS demapper complexity over the uniform QAM demapper complexity
      • Also report if 1D-NUC or 2D-NUC is used, # of spatial layers, and the receiver type (e.g., LMMSE or rML)
    • Also need to report the assumption on complexity counting, e.g, fixed point assumed or floating point assumed
  • The storage requirement for storing all the constellations corresponding to the MCS indices in the MCS table, as a function of precision of constellation point storage
  • Processing delay and parallelism/serialism, if applicable, and their impact to throughput

Note: the complexity is represented by the numbers of comparison, addition/subtraction, and multiplication/division operations, normalized by the number of information bits. Note: For complexity as a function of SE point, will add a column in the already agreed performance reporting table. Note: For complexity/storage not as a function of SE point, will add a row in the already agreed performance reporting table. Note: Spec impact will be separately evaluated, include BICM, scrambling, etc

Note: For 4K uniform QAM DL and 1K uniform QAM UL link level performance study, the following format can be used for performance reporting.

• For assumed TX/RX EVM, before we receive any concrete numbers from RAN4, companies can provide their assumptions. One example can be 6dB tighter than the EVM of 1K QAM for DL and 256QAM for UL.
• Other parameters include: Channel estimation assumption (genie or realistic), channel configurations (AWGN, SISO, SIMO, MIMO and TX/RX antenna configurations, channel types, number of spatial layers,), assumed residual freq offset range, number of allocated RBs, etc
• Two highest MCS points in DL 1K QAM and UL 256QAM in NR added in the table for comparison.
• This is preliminary result and not intended for TR

Note: For high order uniform QAM for DL 4K QAM and UL 1K QAM, to provide the UPT with the high order QAM (DL 4K QAM and UL 1K QAM) over the UPT without the high order QAM under the assumed deployment scenario.

To evaluate the proposal to allow a single spectrum efficiency point to be supported by multiple MCS entries (with different modulation order and coding rate combinations with uniform QAM or with different shaping parameters, coding rate, and constellation size combinations for PS and different coding rate and constellation combinations for GS).

• When providing results, to provide the following information
  • Details on the design of MCS table with overlapping MCS entries and expected size of MCS table, including performance comparison of designs with the same expected size of MCS table
  • Performance benefit under different channel and rank assumptions
    • As baseline, provide performance with legacy MCS table up to 256 QAM
      • Can additionally provide performance with legacy MCS table up to 1K QAM
    • For PS/GS, provide performance allowing each SE point to be mapped to one or more MCS entries
      • For PS/GS, provide performance allowing each SE point to be mapped to only one MCS entry (from the set of one or more MCS entries)
    • For uniform QAM, provide performance allowing each SE point to be mapped to one or more MCS entries
  • MCS selection mechanism across multiple MCS corresponding to the same spectrum efficiency.
    • If UE feedback is needed for gNB to select between multiple MCS entries corresponding to the same SE, provide details on what is to be fed back
• FFS: How different MPR for different modulation order is captured in the simulation for uplink
• FFS: How different EVM for different modulation order is captured in the simulation
• For the purpose of this study, the same set of SE points as in legacy uniform QAM table will be used as starting point.
• When reporting performance, also report other assumptions, including channel type (AWGN, SISO, SIMO, MIMO) and antenna configuration, number of spatial layers, number of RB allocated, TB size, shaping algorithm used (including block length), freq domain interleaver applied or not, receiver assumption, precoding assumption, realistic channel estimation, etc
• To propose how to align shaping parameters or how to align coding rate for facilitating comparison.

For PS/GS fixed MCS performance reporting for 10% BLER (other target x% BLER can also be reported), adopt the following format for simulation as a starting point for result reporting.

Note: For NR MCS reference, since NR has multiple MCS tables, it is not enough to provide the MCS index. Instead, need to provide the (modulation order, coding rate) pair for the simulated SE Note: For SE point specific parameters:

• For GS, this can be a pointer to the constellation used for this SE point
• For PS, this can be a constellation size, coding rate and shaping parameter used for this SE point Note: Other metrics (at least complexity) will be merged into the same table with other columns, if details of the metrics are agreeable. Note: For AMC study, if possible, we can use the same table format

For PS/GS fixed MCS performance reporting for 10% BLER (other target x% BLER can also be reported), adopt the following format for simulation as a starting point for result reporting.

Note: For NR MCS reference, since NR has multiple MCS tables, it is not enough to provide the MCS index. Instead, need to provide the (modulation order, coding rate) pair for the simulated SE Note: For SE point specific parameters:

• For GS, this can be a pointer to the constellation used for this SE point
• For PS, this can be a constellation size, coding rate and shaping parameter used for this SE point Note: Other metrics (at least complexity) will be merged into the same table with other columns, if details of the metrics are agreeable. Note: For AMC study, if possible, we can use the same table format

For DFT-s-OFDM, further study how/whether Net Gain over uniform QAM can be achieved by PS/GS.

R1-2601514 Session Notes of AI 10.3.2 Ad-Hoc Chair (Ericsson) Session notes are endorsed and incorporated the session notes below.

The draft LS in R1-2601663 is endorsed. The final LS in R1-2601664 is endorsed.

For GS, potential impact to the TX/RX chain functionality blocks compared to NR are identified as follows:

• TX chain
  • Mapper
    • Bit to constellation symbol mapping
    • Modulation symbol generation
• RX chain
  • Demodulation of received symbols
  • Demapper

Companies are encouraged to identify potential impact on TX, RX chains if both NUC and uniform QAM are supported. Companies are encouraged to provide design details for the modification needed for above functionalities. Companies are encouraged to explain the reason if a functionality block is not impacted.

Send LS to RAN4 to kindly provide the EVM and MPR values for DL 4K uniform QAM without shaping and UL 1K uniform QAM without shaping.

For PS, potential impact to the TX/RX chain functionality blocks are identified as follows:

• TX chain
  • (Modified) TBS calculation
  • (Modified) CB segmentation
  • (New) DM functionalities
    • Bit splitting: Split to shaped bits and unshaped bits
    • DM
    • Bit concatenation/multiplexing: Concatenate/multiplex DM output and unshaped bits
  • (Modified) Bit interleaver and Bit selection
  • (Modified) Scrambling: shaped bits should not be scrambled to keep the target distribution
  • (Modified) Modulation: Power normalization needed for shaped constellation
• RX chain
  • (Modified) TBS calculation
  • (New) DM functionalities
    • Bit splitting: Split to shaped bits and unshaped bits
    • DDM
    • Bit concatenation/demultiplexing: Concatenate/demultiplex DDM output with unshaped bits
  • (Modified) Bit de-interleaver and Bit selection
  • (Modified) Descrambling:
  • (Modified) Demodulation: Prior probability used in demodulation
  • (Modified) CB concatenation

Companies are encouraged to provide design details for the modification needed for above functionalities. Companies are encouraged to explain the reason if a functionality block is not impacted.

For the study of introducing DL 4K uniform QAM and UL 1K uniform QAM, focus on the following use cases:

• FWA deployed outdoor
• FWA deployed indoor
• FFS: In-door hot-spot.

For SLS of DL 4K uniform QAM and UL 1K uniform QAM for CPE under FWA scenarios, assume the following:

• For layout
  • Dense Urban
  • UMA
• FFS: Other SLS parameters

Parameters affect the PS complexity/storage/latency and BLER and throughput performance trade-off are at least:

• DM output length in unit of I/Q symbol,
• /# of bits shaped per I/Q,
• /# of shaping parameters
• /# of DM blocks needed to support the target throughput of 6GR
• DM algorithm and bit-width of variables in the DM algorithm
• Impact of mismatch on the quantization bitwidth between DM and DDM.

Parameters affect the GS complexity/storage/latency and BLER and throughput performance trade-off are at least:

• 1D-NUC or 2D-NUC
• Constellations and # of constellations
• Bitwidth for describing the constellation
• Bits to constellation mapping

The values of these parameters used in evaluations shall be submitted together with performance results. Companies are encouraged to provide evaluations for different combination parameters to study different performance and complexity/storage/latency trade-offs to provide proper assessment including feasibility.

Study baseline BS setting(s) for evaluating 6G BS EE improvement/impact, considering NR features and 6G BS reference configuration(s) Study baseline UE setting(s) for evaluating 6G UE EE improvement/impact, considering NR features and 6G UE reference configuration(s)Agreement

Study metric(s) for UE energy efficiency. Study metric(s) for BS energy efficiency.

Study how/whether to reuse or update the power model in TR 38.864 for evaluating BS power consumption for 6G BS.

Study how to reuse and update reference configurations in TR 38.864 for 6G BS.

Study reference configurations and power consumption model for 6G UE, considering but not restricted to the following:

• TR 38.840 (UEPS), TR38.875 (RedCap), TR38.865 (eRedCap), and TR38.869 (LP-WUS/WUR) for reference configurations
• TR 38.840 (UEPS), TR38.875 (RedCap), and TR38.869 (LP-WUS/WUR) for power consumption models

Study whether/how to further update the BS model considering the following aspects, e.g.,

• Whether to downselect between Cat.1 and Cat.2,
• Updates of parameter values (including defining a new Cat),
• Updates of power scaling, power states (including additional PSs)
• Etc. Note: The defined BS power models does not preclude use case-specific enhancements regarding, e.g., multi-TRP, SBFD, multi-carrier etc

Study and evaluate NW energy savings and the impact on UE performance and user experience with respect to 20ms and longer periodicities of sync signal(s) at least for initial access with the following consideration, but not limited to: BS assumptions:

• Cell-common signaling (e.g., sync signal(s), broadcast PDCCH, SIB-1, SIB, paging, PRACH), e.g.,
  • Clustered provisioning of different cell-common signaling,
  • On-demand provisioning of different cell-common signaling,

• UE-specific signaling (for low, light, medium loads), e.g.,

  • Clustered provisioning with cell-common signaling,
  • Unclustered provisioning with cell-common signaling, UE impact:

• Cell search complexity and latency, including frequency search latency,

• UE Power consumption,
• Sync signal detection, coverage and tracking performance,
• RRM, mobility,
• Beam management,
• Other properties are not precluded,
• Improvements to address identified impact, e.g.,
  • Additional sync signal needs,
  • Adaptation of sync signal transmission periodicity,
  • Sparser synch raster.

Study and evaluate on-demand and/or periodic SIB-1 transmission with respect to

• NW energy savings potential and UE power consumption impact,
• SIB-1 acquisition delay,
• NW and UE complexity,
• Coverage,
• Applicable deployment scenarios, e.g.,:
  • Standalone cell/carrier,
  • Multiple TRPs/cells/carriers.

For evaluation purposes, study extending NR UE power consumption model for UE operation with DL WUS of OFDM-based sequence, regarding the following aspects:

• Power state(s), sleep and non-sleep, and corresponding characteristics and power value(s)
• Transition energy and time for each of sleep state(s)
• Companies to report the assumption(s) for achieving the proposed power value(s), e.g., time/frequency domain detection, noise figure assumption(s), synchronization assumption(s), BW/antenna assumption(s), etc.

Apply the following evaluation methodology framework for Quantitative analysis,

• For NW unloaded/empty load case or UE idle/inactive mode:
  • For energy saving: analytical calculation
  • For performance impact: analytical calculation, LLS
• For loaded cases and connected-mode UEs
  • For energy saving: SLS
  • For performance impact: LLS, SLS

At least the following NR metrics,

• Network energy saving gain relative to baseline for BS
• UE energy saving gain relative to baseline for UE
• Impact to UPT (User-Perceived Throughput), if applicable, as well as the metrics
• Impact to latency, if applicable
• Impact to QoS/delay budget satisfaction rate, if applicable are used for 6G energy efficiency evaluation.

For evaluation purposes, expand the existing BS power model reference configuration with a set for ~7 GHz operation with the following parameters:

Note: Bracketed values to be confirmed. Other values are not precluded. The above configuration has no implication on supported BW, SCS for 6GR.

For 6GR energy efficiency evaluation purposes, reuse the existing UE power consumption model FR1 and FR2 reference configurations in TR 38.840 for operation up to around 7GHz and within 24.25 GHz – 52.6 GHz, respectively.

• Scaling rules can be updated, including additional rule(s) for scaling UE power consumption, and including around 7GHz specific update
  • FFS: details.
• Power value and transition time update, if necessary, including around 7GHz specific update
• No implication on supported BW, SCS, modulation and antenna setting for 6GR
• Revisit if SCS for around 7GHz is different with respect to the reference configuration

Study and evaluate DL WUS of OFDM based sequence and corresponding mechanisms for 6GR EE improvement, regarding at least the following aspects:

• Coverage target for DL WUS (e.g., same as PDCCH, common sync signal, or other)
• Measurements and/or synchronization.
• System overhead and network energy consumption/UE energy saving for UE operation with the DL WUS.
• RRC states
• Other functionalities

For evaluation purposes, study extending NR UE power consumption scaling w.r.t. at least BW and/or antenna setting, considering at least the different characteristics in RF/BB power consumption and static/dynamic power consumption.

Study and evaluate multi-carrier/cells/TRPs mechanisms for 6GR NES, considering, e.g.,:

• Sync signal-less carriers/cells/TRPs for at least intra-band and collocated inter-band multi-carrier/cell/TRPs, including potential extensions to additional deployments and scenarios,
• RRC states,
• UE energy consumption and complexity,
• Other mechanisms/aspects/signals/channels are not precluded.

Study and evaluate on-demand sync signal(s) mechanisms for 6GR energy efficiency, considering, e.g.,:

• On-demand sync signal(s) for single cell/carrier, multi-carrier/cell, multi-TRP,
• Network-triggered and UE-triggered on-demand sync signal(s),
• Idle and/or connected modes,
• Other mechanisms/aspects/signals/channels are not precluded.

6GHOW NOTE: SOME FORMULAS IN THESE AGREEMENT IS INCORRECT. NEEDS TO BE FIXED.

For the UE power model, assume:

• \(\Gamma_B\): The ratio of active bandwidth to the bandwidth of reference configuration (\(B_{ref}\))
• \(\Gamma_{Tput}\): The ratio of the maximum schedulable PDSCH throughput w.r.t. that of current active bandwidth,
  • FFS: How the maximum schedulable PDSCH throughput ratio \(\Gamma_{Tput}\) is ensured by NW
  • Companies to report the assumed technique to achieve the reduction in the maximum schedulable PDSCH throughput. Current techniques include at least: reduced scheduled maximum bandwidth, reduced scheduled maximum rank, and reduced scheduled duty cycle.
  • No antenna adaptation is assumed simultaneously
• For evaluation purpose,
  • Company to report \(d\) value for the evaluation
• \(D > d\) is another relaxed adaptation delay. Company to report \(D\) value for the evaluation.
• The adaptation delay is applied when adapting \(\Gamma_B\) and/or \(\Gamma_{Tput}\)
• Companies to report \(S_{offset}\) from one of the two options:
  • Option 1: \(S_{offset} = 0 / Z_1 / Z_2\) for Maximum BW (MaxBW) assumed for the evaluation \(\in {100 \text{ MHz}, 200 \text{ MHz}, 400 \text{ MHz}}\)
  • Option 2: \(S_{offset} = 0\)
• Note: No implication on which configuration(s) and what adaptation(s) to be supported by 6GR

Include the following scaling for DL bandwidth adaptation for 6GR UE power consumption model:

• For PDCCH+PDSCH:

• For PDCCH-only:

• Other DL signal/channel processing:

• Not applicable to sleep states and EE processing state
• Further consolidate the table for cases of a same scaling factor

• FFS: Scaling factor for other value(s) of \(\Gamma_{Tput}\)

• For UL bandwidth adaptation,

  • No scaling for FR1 and around 7GHz
  • In case scaling is needed for FR2 (including 24.25 GHz – 52.6 GHz), companies can report the assumed scaling factor
• Note: The term “bandwidth” used here is reusing the definition as 5GNR, which will be updated according to 6GR discussion
• Note: All columns and rows of the above tables will be further checked, and corresponding values will be checked and confirmed/extended.

Add the following as one of reference configurations for BS power consumption

**Pending agreement in 11.2 whether to evaluate 15 GHz.

Include the following non-sleep states as 6G UE power consumption model.

• Other power state(s) is not precluded
• FFS: Configuration and relative power value(s) for EE processing in FR2 (including 24.25 GHz – 52.6 GHz)
• Note: Pending agreement in 11.2 whether to evaluate 15 GHz

Include the following sleep states for 6G UE power consumption model. Note: Ultra Deep Sleep is not intended for connected mode

• Y value equals to 0.1, if EE processing is assumed for the evaluation; zero, otherwise.

• Immediate transition is assumed for power saving study purpose from or to a non-sleep state ** Ramp-up time is no less than half of the total transition time *** Time for sync/re-sync is not included

For evaluation purposes and relative comparison over different candidate energy saving schemes for 6GR, define the following baseline power saving configurations for UE for evaluation purpose for FR1 (including around 7GHz):

• 5G NR I-DRX (1.28s cycle) for idle mode
  • Group paging rate (for a PO): TBD
• 5G NR C-DRX settings of (cycle, on-duration timer, inactivity timer) are assumed for the following 6GR traffic models for connected mode:
  • VoIP: (40 ms, 8 ms, 10 ms)
  • FTP3: (160 ms, 8 ms, 100 ms)
  • Instant message: (320 ms, 8 ms, 100 ms)
  • XR: (16 ms, 10 ms, 4 ms)
• Companies can evaluate and report other traffic(s) and/or configuration(s) with justification Note: The corresponding evaluation is not intended for energy efficiency comparison with 5G/NR.

Total DL power level for Set 4 is 56 dBm. The following relative power levels applies to Set 4 for a CAT 1 BS and a CAT 2 BS:

Include the following UL long PUCCH/PUSCH/PRACH power values in the UE power model: Note: UE reference configuration is 1TX chain

The following relative transition energies are adopted for BS models CAT1 and CAT2, if supported, for BS reference configuration Set 4:

FFS: whether delta value is needed due to larger number of TXRU

Include the following bandwidth scaling table for micro sleep in the 6GR UE power model:

• Subject to adaptation delay \(= T_{min}\), no larger than NR BWP switch delay (Type 2)
  • \(S_{MaxBW}\) can be zero, or the other option: 0.3 and 1.0 for \(MaxBW=200\%\) and 400%, respectively. Company to report which option assumed in their evaluation.

• If adaptation delay \([5ms] \le T \le [10 \text{ ms}]\) is allowed, scaling factor for \(\Gamma_B \le 25\%\) can be 0.8 or the other option: 0.6. Company to report which option is assumed in their evaluation.

The following scaling w.r.t. DL antenna number is also applied to micro sleep:

FFS: Whether the above scaling rules is applicable to light sleep

Update PDCCH-only bandwidth scaling table in the 6GR UE power model with the following values:

• Subject to adaptation delay \(= T_{min}\), no larger than NR BWP switch delay (Type 2)
  • \(S_{MaxBW}\) can be zero, or the other option: 0.3 and 1.0 for \(MaxBW=200\%\) and 400%, respectively. Company to report which option assumed in their evaluation.

• If adaptation delay \([5ms] \le T \le [10 \text{ ms}]\) is allowed, scaling factor for \(\Gamma_B \le 25\%\) can be 0.8 or the other option: 0.6. Company to report which option is assumed in their evaluation.

Update PDCCH+PDSCH bandwidth scaling table in the 6GR UE power model with the following values

• Subject to adaptation delay \(= T_{min}\) no larger than NR BWP switch delay (Type 2)
  • \(S_{MaxBW}\) can be zero, or the other option: 0.15 and 0.5 for \(MaxBW=200\%\) and 400%, respectively. Company to report which option is assumed in their evaluation.

• If adaptation delay \([5ms] \le T \le [10 \text{ ms}]\) is allowed, scaling factor for \(\Gamma_B \le 25\%\) can be 0.4 or the other option: 0.3. Company to report which option is assumed in their evaluation.

RAN1 to further define reference configuration for UEs with 20MHz or smaller bandwidth.

• FFS: Same scaling rules (without SMaxBW) are applied to both reference configurations.
• FFS: Reuse/extend the reference configuration and power values from NR TR 38.875

Include the following DL antenna scaling factors in 6GR UE power consumption model: Note: Applicable for FR1 (including around 7GHz) and FR2 (including 24.25 GHz - 52.6 GHz), where, for FR2, number of DL antenna assumed is up to [4] FFS: Scaling for 8Rx

Adopt the following cross-slot scheduling scaling for evaluation for 6GR UE power consumption model:

For NES evaluation purposes and relative comparison of different candidate energy saving schemes for 6GR, define the following baseline network configurations

• SSB with 20 ms periodicity, at least for single cell
• SIB1, if available, company to report assumed periodicity from {20 ms, 160 ms}

• RO, if available, with 10/20 ms periodicity Furthermore, to assist comparisons

• Companies to report \(s_{tx}\), \(s_{rx}\) and \(s_{f}\) values for BS processing of the above signal(s)/channel(s)

• Companies to report the average network load in %
• Companies can evaluate and report other configuration(s) with justification Note: The corresponding evaluation is not intended for energy efficiency comparison with 5G/NR.

IF a BS model CAT 2.1 (2-plus) is introduced, it has the following transition energy characteristics for Set 1-3. FFS: Set 4:

The following transition times are adopted for BS models CAT1 and CAT2, if supported, for BS reference configuration Set 4:

FFS: whether delta value is needed due to larger number of TXRU

Adopt the following UL scaling for 6GR UE power consumption model:

FFS: Scaling for different symbol numbers

For CA reception based on one RF chain, power scaling factor is same as that of bandwidth scaling with the same total bandwidth.

• Scaling on power consumption assuming all CCs are activated, and no implication on carrier activation/deactivation delay.
• FFS: Separated scaling for slots with PDCCH-only
• FFS: Scaling for CA reception based on more than one RF receive chains.

In EE evaluations, to adapt a BS's number of active antenna TXRU to \({1/2 \text{ and } 1/4}\) of the total number of antenna TXRU and to adapt the associated circuitry. FFS:

• Power model for BS
  • Power states
  • State transitions
  • Application of scaling of BS for static and/or dynamic parts
• Coverage, throughput aspects in UL and DL
• BS models and reference configuration compatibility
• Applicable signals and channels, related procedures and RRC states
• Latency aspects
• UE power consumption and complexity aspects
• BS complexity
• Use cases (e.g., UE-initiated on-demand SIB1)

BS model CAT 2.1 is introduced. All BS models CAT 1, CAT 2 and CAT 2.1 may be used for evaluations in 6GR SI.

For EE Processing 2-RX in FR1, adopt the following values:

If EE processing reception time is X symbols within a slot, the power value is scaled by \((X/14)\).

• FFS: whether to scale additional energy overhead
• FFS: Values for 10MHz BW

Study link direction determination for dynamic TDD, considering at least

• UE PDCCH monitoring efforts and power consumption
• Signaling overhead
• Scheduling flexibility
• Avoid duplicated functionalities
• Collision handling

6GR shall at least be capable of configuring the same TDD slot configurations as TDD slot configurations deployed in 5G NR.

For the RAN1 study of "Re-use of existing 5G mid-band (3.5GHz) site grid for 6G deployments in at least around 7 GHz and targeting comparable coverage to 5G mid-band", the following assumptions are assumed for link budget template candidates 1 for signals/channels in around 7GHz

Note: Companies to provide updated link budget results before April 3rd, to be triggered by email thread (March 30th).

For the RAN1 study of "Re-use of existing 5G mid-band (~3.5GHz) site grid for 6G deployments in at least around 7 GHz and targeting comparable coverage to 5G mid-band", the following assumptions are assumed for link budget template candidates 1 for Msg3 PUSCH in 5G mid-band

For the RAN1 study of "Re-use of existing 5G mid-band (~3.5 GHz) site grid for 6G deployments in at least around 7 GHz and targeting comparable to same coverage to 5G mid-band",

• For the link budget evaluation for coverage gap identification in around 7 GHz
  • For initial access, Rel-15 NR signals/channels during initial access are used for identifying the gap of individual signal/channel compared to Rel-15 NR msg3 in 5G mid-band, respectively Note: The candidate coverage enhancement techniques will be separately discussed.

For 6GR spectrum aggregation operation, study the following methods and their associated application scenarios:

• Method 1: Multiple physical carriers can be aggregated into single "Gothia cell"
  • Note: the term 'Gothia cell' is for RAN1 discussion purposes, and whether/how to specify the feature / refer to the feature is separate RAN1 discussion.
• Method 2: "Carrier aggregation" where multiple physical carriers can be aggregated into separate cells For both methods, study them under idle mode and connected mode, and study their pros and cons at both NW and UE side

For the 6GR smallest maximum UE bandwidth for at least one lower-tier device, RAN1 to further study the following alternatives:

• Alt 1: 20MHz RF and BB bandwidth for both UL and DL with 15kHz SCS for FDD, and with 30kHz SCS for TDD
• Alt 2: 5MHz RF and BB bandwidth for FDD with 15kHz SCS for both UL and DL, 10 or 20 MHz RF and BB bandwidth for TDD with 30kHz SCS for both UL and DL
• Alt 3: 20MHz RF bandwidth for both UL and DL with 15kHz SCS for FDD, and with 30kHz SCS for TDD
  • Narrower bandwidth for BB in UL, and/or Narrower bandwidth for BB in DL

Basic 6GR sync signal structure is defined as 6GR SSB, which consists of primary synchronization signal(s), secondary synchronization signal(s) and physical broadcast channel(s)

• FFS: Other types of 6GR sync signal/channel structure or reference signal and their structure

For initial access and mobility in 6GR, study the following deployment scenarios

• Single beam and multi-beam based deployments
• Single TRP and multi-TRP based deployments
• Single carrier and multi-carrier deployments
• Other deployment scenarios

For 6GR paging transmission/reception, study at least the following aspects:

• Study paging transmission scheme(s) to facilitate network energy savings
• Study paging scheme(s) to facilitate UE energy savings
• Study necessity of paging capacity enhancement
• Study necessity of paging coverage enhancement

Study 6GR signals, channels and procedures for initial access and idle mobility, considering at least

• Cell/Initial Cell search and cell ID identification
• Time/frequency synchronization/tracking
• Beam measurement
• System information acquisition
• Whether TRP is transparent/non-transparent to UE during above procedures

Study 6GR signals, channels and procedures for idle mobility, considering at least

• Cell search and cell ID identification
• Time/frequency synchronization/tracking
• Beam measurement
• System information acquisition
• Whether TRP is transparent/non-transparent to UE during above procedures

For 6GR measurements in initial access and for mobility, study measurement resource, measurement quantity, measurement functionality and measurement procedure, at least including:

• L1 measurements
• Cell-level/[TRP-level] and beam-level measurement

Study beam operation during 6GR initial access, including:

• Beam acquisition during initial access
• Beam report/refinement during initial access
• Feasibility and performance of spatial/temporal beam prediction during initial access

Adopt the following link level simulation assumption for random access evaluations: Link Level Assumption Parameters for Random Access

Note: additional parameter tables are evaluation parameters specific to the evaluation of PRACH or Msg 3 that would override the general link level assumption parameters for random access if fields were duplicate

Additional Parameters for PRACH Evaluations

NOTE 1: The CDL table is translated so that the strongest cluster's AoD and AoA occur at a random angle for both the antenna panels of TRP and UE in the local coordinate systems. ZoD and ZoA is assumed to be unchanged. The value of the random angle is selected to be uniformly distributed from +30 to -30 degree. The random value is chosen independently for both AoD and AoA. CDL angle scaling is based on Clause 7.7.5.1 of TR38.901 v19.1.0.

Additional Parameters for PUSCH of Msg.3

Study random access framework with the following aspects:

• Enablement of energy efficient random access procedures (supporting SID objective 1b);
  • Including both network and UE power saving
• Coverage improvement (supporting SID objective 1d);
• Support of random access for diverse device types and capabilities (supporting SID objective 1g);
• System performance improvement from overhead reduction, simplification of signaling/configurations (supporting SID objective 1k);
• Additionally consider following aspects
  • random access latency;
  • capacity
  • detection reliability;
  • high speed mobility;
• Note: Other aspects identified during future discussions are not excluded.

The following scenarios and assumptions beyond single carrier/TRP are considered for the study of above random access framework:

• NTN
• SBFD
• multi-carrier
• multi-TRP
• Note: whether/how to support one or more of the scenarios/assumptions, including whether any special handling or functionality needs to be introduced in support of the scenarios/assumptions is part of the study.

For basic initial access procedures considering UEs with different bandwidth capabilities, study the determination of the frequency location(s) and bandwidth(s) for UE to transmit uplink signals/channels during initial access and idle mode:

• FFS: Relation to the DL frequency location(s) and bandwidth(s) in TDD
• FFS: Relation to frequency location of 6G synchronization signals

For basic initial access procedures considering UEs with different bandwidth capabilities, study the determination of frequency location(s) and bandwidth(s) for UE to receive/monitor the downlink signals/channels during initial access and idle mode for at least the following:

• System information
• Random access procedure
• Paging
• FFS: Relation to frequency location of 6G synchronization signals

6G PDCCH study will consider at least the following concepts:

• CORESET
• CCE
• REG as the minimum resource unit
• REG bundle
• CCE to REG mapping
• Search Space
• PDCCH candidate and CCE aggregation levels
• Blind decoding
• DMRS for PDCCH
• Hash function FFS the details of the concepts above FFS the relation among CCE, REG, REG bundle, and PDCCH candidates

• REG consists of multiple REs.
• CCE consists of a set of REGs.
• PDCCH candidate corresponds to one or multiple CCE(s).
• REG bundle consists of a set of REGs.

For 6G PDCCH study, consider at least the following aspects:

• Control-channel blocking
• Coverage
• UE energy consumption
• UE complexity
• Coexistence of different device types
• Reliability of PDCCH transmission
• NW complexity and scheduling
• Resource efficiency and overhead
• Latency
• MRSS

Study PDSCH and RS for PDSCH based on the following SLS EVM assumptions

• Note: Additional EVM assumptions for AI/ML based DMRS overhead reduction can be further discussed.
• Note: EVM assumption for HST scenarios will be treated separately

Study PDSCH and RS for PDSCH based on the following LLS EVM assumptions Note: Additional EVM assumptions for AI/ML based DMRS overhead reduction can be further discussed.

Study the following options regarding the spec impact on the maximum number of orthogonal DMRS ports for PDSCH

• Option 1: Up to 24 orthogonal DMRS ports
• Option 2: Up to 32 orthogonal DMRS ports
• Option 3: Up to 48 orthogonal DMRS ports
• Option 4: Up to 64 orthogonal DMRS ports
• Option 5: Up to 96 orthogonal DMRS ports Study Non-orthogonal DMRS on top of the options above to achieve the target number of MIMO layers (from network side) Note: To provide link/system level simulation results for this study

Study the PT-RS for PDSCH including at least the following aspects

• The necessity of PT-RS in different bands

If 6GR supports to define uplink control information (UCI) in Layer 1, study how to define the method(s) to convey the UCI over physical channel, at least (but not limited to) the following aspects:

• Option 1: Define a UCI dedicated physical channel, i.e., Physical Uplink Control Channel (PUCCH)
• Option 2: UCI carried on PUSCH
• Option 3: Other method(s) Note: Whether UCI is carried in L1 or L2 to be discussed in agenda item 10.5.3.1 and 10.5.4.3

Support following table as the basic assumption of SLS for evaluation of PUSCH transmission scheme. Note: Additional EVM assumptions for AI/ML based evaluation can be further discussed.

Note: EVM assumption for HST scenarios will be treated separately

Study of UE-initiated/event-driven beam management (UEIBM) mechanisms for 6GR, covering at least the following aspects:

• Event definition and the corresponding target use case.
• UL transmission, UE's behavior, and procedure for a triggered event, including the necessity and design of the associated network's response. Note 1: Both AI/ML and non-AI/ML related mechanism(s) for the above can be further studied. Note 2: UE-initiated/event-driven CSI reporting is not discussed in this agenda.

Study TCI/QCL-related aspects, e.g., definition of QCL/TCI-state, QCL property/chain On beam management for DL and UL of 6GR, at least the following aspects should be studied:

• Beam measurement(prediction)/report/indication within a same TRP, i.e., single-TRP, in a cell/carrier;
• Beam measurement(prediction)/report/indication among different TRPs, i.e., multi-TRP, in a cell/carrier;
• Beam measurement(prediction)/report/indication among different cells/carrier, i.e., inter-cell/carrier Note: Both AI/ML and non-AI/ML related mechanism(s) for the above can be further studied. Note-1: Which multi-TRP transmission scheme for study will be discussed under other agenda.

Regarding link-level evaluation of 6GR beam management, in RAN1#124b, to use the following template as starting point for collecting related parameters.

Regarding system-level evaluation of 6GR beam management, in RAN1#124b,, to use the following template as starting point for collecting related parameters.

Adopt the following table as the basic assumption of SLS for evaluation of DL-based CSI acquisition. Table General Assumption

Note: EVM assumption for HST scenarios will be treated separately

Adopt the following table as the assumptions for LLS for DL based CSI acquisition.

Note: additional necessary evaluation assumptions for CSI feedback without channel coding, e.g. JSCM, JSCC, are to be discussed separately.

For evaluation of multi-TRP under single layer deployment,

• Same antenna configuration and Tx power is assumed across TRPs. Note: other multi-TRP scenarios are not precluded Note: TRP assumption for HST scenarios will be treated separately

6GHOW NOTE: IMAGES ARE NOT YET SUPPORTED.

Table Multi-TRP scenarios

To evaluate the performance of non-AI/AI based CSI prediction at UE side or NW side, consider:

• Intermediate metric: Subband/RB-level per layer eigenvector/precoder SGCS, NMSE calculated between predicted channel and ground-truth channel
• To compare NW side prediction with UE side prediction, the impact of CSI reporting shall be considered
• FFS: other intermediate KPI
• System-level metric: UPT (mean, 5%tile), CSI reporting overhead
  • CSI-RS overhead is considered in UPT calculation
  • Companies to report how to calculate CSI report overhead and CSI report mechanism
• Link-level metric: BLER/SE/throughput
• Complexity metrics: FLOPs/M

• For AI based prediction: number of parameters/M Baseline:

• Baseline#1: CSI measurement/report based on full CSI-RS ports/density

• Baseline#2:
  • Sample & Hold for frequency domain, companies to report how Sample & Hold is performed
  • FFS for spatial domain
• Other baselines are not precluded
• Companies to report sampling ratio and sampling pattern
  • Definition of sampling ratio and sampling pattern subjects to detailed domain for reduction
• FFS calibration error for BS antenna arrays other than 4TXRUS
• Companies to report the assumptions for spatial consistency modelling, if applicable The above is at least applicable to frequency/spatial domain prediction. Companies to report other sub-use case specific parameters, if applicable

For the study on AI-based CSI compression, the following is considered:

• KPI: SGCS (Intermediate), PAPR (when applicable), throughput vs CSI feedback overhead, DL throughput/BLER vs SNR
  • CSI feedback overhead: the number of UL resource elements
  • Companies to report values
  • Note: the overhead of DMRS (if applicable) is reported by company
• Model/computation complexity at UE side (when applicable): FLOPS/M
• Model/computation complexity at NW side: FLOPS/M
• Model size: Number of parameters/M
• Overhead of downloadable basis/codebook/matrix for compression (when applicable): size of parameters (Mbytes) and updated frequency
• Overhead/complexity/impact of inter-vendor collaboration (when applicable)
• Benchmark:
  • Rel-20 NR SSCC approach with 2-sided model (companies report the inter-vender collaboration assumption) and/or NR eType II feedback and/or NR Type I feedback
• Link adaptation is considered
• Evaluation assumptions follow general EVM.
  • For uplink transmission (e.g., UL channel model, UL SNR range, UL channel estimation, UL noise + interference estimation) are reported by companies
  • Companies to report the considered non-ideal factors, if assumed, e.g., Tx PA nonlinearity & RF impairment, UL/DL channel estimation errors, uplink interference pattern
  • Companies to report the bitwidth/shape of the modulation constellation
• Note: the above assumptions can also apply for non-AI schemes, i.e., without training/inference/data collection/projection, when applicable.
• Note: RAN4 related impacts, e.g., EVM, ACLR, SEM, are to be considered

Study at least the following aspects of SRS for uplink and downlink CSI acquisition:

• Efficient support of larger channel bandwidth
• Capacity enhancements
• Coverage enhancements
• Efficient resource utilization
• Dynamic/flexible adaptation of SRS parameters
• mTRP transmission/reception, FWA, HST and other high mobility scenarios
• Interference mitigation
• Energy efficiency

For UL-based CSI acquisition, study at least the following SRS usages:

• CSI acquisition for UL transmission
• CSI acquisition for DL transmission
• Beam management
• Other usages are not precluded

Adopt the assumptions in the following table as the basic assumption of SLS for evaluation of UL-based CSI acquisition.

Table: SLS assumption for DL CSI and UL CSI

Note: EVM assumption for HST scenarios will be treated separately

Adopt the assumptions in the following table as the basic assumption of LLS for evaluation of UL-based DL CSI acquisition.

For LLS for UL-based UL CSI acquisition, reuse the LLS assumption for uplink transmission schemes under AI 10.5.2.3.

Considering at least the following aspects for the tracking RS design for finer time/frequency tracking Set 1 (KPI):

• Tracking performance
• Overhead
• Energy efficiency for NW&UE
• UE-side complexity
• Accuracy of QCL parameter(s) estimation

For joint DL and UL based DL CSI acquisition, reuse the evaluation methodology for LLS and SLS agreed in DL CSI / UL CSI agendas. Note: Specific necessary aspects on joint operation will be discussed.

Adopt the following simulation assumptions for tracking RS evaluation Table 1: LLS assumptions for T/F tracking

Study joint DL and UL based CSI acquisition in TDD system.

Study and evaluate at least the following aspects on RS for finer time/frequency tracking for both connected and idle modes

• Frequency domain factors
  • Bandwidth
  • Frequency domain density
• Time domain factors
  • Time domain behavior: e.g., periodic, semi-persistent, aperiodic, on-demand
  • Time domain density: e.g., periodicity, number of slots, number of symbols and interval between symbols in a slot, etc.
• Spatial domain factors
• FFS: Details

Study single-stage and two-stage DCI.

RAN1 to study what functions are indicated by DCI for 6GR. RAN1 to study how functions are indicated by DCI for 6GR.

Study the adaptation of PDCCH monitoring in 6GR for UE energy efficiency purpose, considering at least:

• Impact on network energy efficiency, as well as complexity and performance
• UE energy saving gain Note: 6GR DL WUS related discussions are handled in AI 10.6.

RAN1 to study L1 signalling framework for time-domain scheduling/resource allocation for downlink and uplink transmission by considering, e.g., within a slot or across the slot boundary, PXSCH repetitions, and other necessary aspects if any.

• Note1: PXSCH corresponds to PDSCH or PUSCH.
• Note2: it doesn't mean those examples are confirmed
• Note3: The descriptions/figures for scheduling options in [R1-2601584, FL summary] is an example for information.

In 6GR, DL and UL HARQ operation designs considers at least the following aspects:

• latency
• reliability
• coverage
• power saving (NW and UE)
• NW complexity
• UE complexity
• diverse services/applications/traffics
• system efficiency/system throughput/user throughput
• feedback efficiency/UL and DL overhead Note: the design of DL and UL HARQ does not necessarily be the same

Study possible HARQ-ACK payload size range.

For discussion purposes,

• Asynchronous HARQ refers to that retransmission(s) occurs in a non pre-determined occasion once the corresponding initial transmission is scheduled.
• Adaptive HARQ refers that the transmission parameters and resources for the retransmission can be adaptively adjusted. For DL and UL in 6GR, support asynchronous and adaptive HARQ operation.

For DL HARQ in 6GR, study both following HARQ-ACK feedback mechanisms

• Mechanism 1: HARQ-ACK information bits are transmitted via L1 signalling
• Mechanism 2: HARQ-ACK information bits are transmitted via higher layer signalling (e.g., MAC CE)

In 6GR, support at least TB level granularity for HARQ-ACK feedback

Study the UE reporting mechanism for requesting uplink resource with considering at least the following aspects:

• Applicable use cases, at least including requesting resource for UL data transmission
• UL transmission latency
• Signalling overhead
• UL resource efficiency
• System capacity
• Network/UE complexity
• other aspects are not precluded

Study the UE-to-UE cross-link interference for 6GR:

• Study and identify the applicable scenarios of UE-to-UE cross-link interference.
• Study the characteristics of UE-to-UE cross-link interference, for example what is the potential interference signal level, what is the potential impact on system operation.
• Study the mechanisms and techniques for handling UE-to-UE cross-link interference, including UE-to-UE cross-link interference measurement and reporting
  • For each candidate mechanism, evaluate and analyze the performance benefit, impact to the system and complexity at BS/UE.
• Study the candidate resource or signals/channels for measuring UE-to-UE cross-link interference. Study and identify the candidate measurement quantities.
• Study the reporting mechanisms

Study BS-to-BS cross-link interference for 6GR:

• Study and identify the applicable scenarios of BS-to-BS cross-link interference.
• Study the characteristics of BS-to-BS cross-link interference, for example what is the potential interference signal level, what is the potential impact to the system operation.
• Study the mechanisms and techniques that can handle the BS-to-BS cross-link interference, including the mechanisms for measuring and identifying the BS-to-BS cross-link interference:
  • For each considered mechanism, evaluate the performance benefit, impact to the system operation, and complexity at BS/UE.
  • The candidate resource for measuring BS-to-BS cross-link interference.
  • The measurement quantities at least for evaluation purpose.

For handling remote interference in 6GR:

• Study the applicable scenarios for remote interference between remote cells due to atmospheric ducting;
• Study the impact of remote interference to the system, including the impact to the uplink reception.
• Study the characteristics of the remote interference.
• Study the candidate mechanisms for measuring/detecting/identifying remote interference, mechanisms to mitigate remote interference.
  • For each candidate mechanism, evaluate the benefits and impact to the system. Note: UE reporting to request uplink resource scheduling is to be discussed under this agenda

Targeting for same coverage as 6G PDCCH in the same band, Study DL WUS coverage by considering at least the following aspects.

• Missed detection rate
• False alarm rate
• RRC state differences
• Different use cases
• Reference configuration of PDCCH

The same SCS is assumed for DL WUS as for the 6GR Sync signals in the same band if sync signals and data channels use the same SCS.

• FFS SCS of DL WUS if sync signals and data channels use different SCS in FR2-1

Study the following aspects of DL WUS design:

• Time or frequency domain sequence definition
• Time and frequency resource allocation
• Multiplexing and coexistence with other signals and channels, including DL WUS
• PAPR and BS/UE processing complexity
• Network overhead and NW&UE energy efficiency
• Other aspects are not precluded

For RRC idle state, study serving cell RRM measurement based on 6GR measurement signal (e.g., at least 6GR sync signal) by EE processing, considering at least:

• UE energy saving gain
• measurement based on EE processing together with DL-WUS monitoring vs measurement based on non-EE processing together with DL-WUS monitoring
• coverage (e.g., achievable SINR/SNR) and accuracy
• Impact on the EE processing complexity

For RRC idle state, study neighboring cell RRM measurement based on 6GR measurement signal (e.g., at least 6GR sync signal) by EE processing, considering at least:

• UE energy saving gain
• measurement based on EE processing together with DL-WUS monitoring vs measurement based on non-EE processing together with DL-WUS monitoring
• Neighboring cell identification, measurement and evaluation
• Neighboring cell number limitation, if any
• Inter-cell interference
• Coverage (e.g., achievable SINR/SNR) and accuracy
• Impact on the EE processing complexity
• FFS the power consumption of neighboring cell RRM measurement based on EE processing Note: It doesn't mean the measurement in EE processing has to be coupled with DL WUS monitoring

Study 6GR DL WUS triggering PDCCH monitoring with and without C-DRX in RRC connected state, considering at least:

• UE energy saving gain
• Impact to UPT, if applicable
• Impact to latency, if applicable
• Impact to QoS/delay budget satisfaction rate, if applicable
• Network overhead/complexity
• Network energy consumption
• Other impacts, if any

Study further necessity and feasibility of UL WUS under the following deployment scenario (DS):

• DS#2: Multi-cell/carrier where UE obtains UL WUS configuration from cell/carrier #1

  • Always-on 6GR synchronization signals with PBCH on cell/carrier #1, and SIB1 on cell/carrier #1
  • DS#2a: Prior to UL WUS, nothing transmitted on cell/carrier #2
  • DS#2b: Prior to UL WUS, always-on 6GR synchronization signals with or without PBCH on cell/carrier #2 but no SIB1 on cell/carrier #2
  • DS#2c: Prior to UL WUS, 6GR synchronization signals with PBCH on cell/carrier #2 and periodic SIB1 on cell/carrier #2 Note: UL WUS can be transmitted to cell/carrier #1 or cell/carrier #2 RAN1 to consider at least

• Coverage target of UL WUS

• How UE acquires synchronization for UL WUS
• How UE decides UL WUS transmission power
• Whether the cells/carriers are in the same band or not
• Whether the cells/carriers are collocated or not
• RRC states for the above scenarios
• Whether the above scenarios can be applicable to multi-TRP deployment scenario Note: The above scenarios are for study purposes only and do not imply 6GR support.

Study further necessity and feasibility of UL WUS under the following deployment scenario (DS):

• DS#1: Standalone cell where UE obtains UL WUS configuration e.g., from standalone cell and/or pre-defined in the specifications

  • DS#1a: Prior to UL WUS, nothing transmitted on the standalone cell
  • DS#1b: Prior to UL WUS, always-on 6GR synchronization signals with or without PBCH on the standalone cell but no SIB1 on the standalone cell
  • DS#1c: Prior to UL WUS, always-on 6GR synchronization signals with PBCH on the standalone cell and periodic SIB1 on the standalone cell Note: UL WUS is transmitted to the standalone cell RAN1 to consider at least

• Coverage target of UL WUS

• How UE acquires synchronization for UL WUS
• How UE decides UL WUS transmission power
• RRC states for the above scenarios Note: The above scenarios can be applicable to multi-TRP deployment scenario if TRP is transparent to UE at the time of UL WUS transmission Note: The above scenarios are for study purposes only and do not imply 6GR support.

RAN1 to use the following terminology when discussing GNSS availability at least for physical layer operation.

• GNSS-based: Refers to the network mode of operation which relies on devices being equipped with a GNSS receiver and the devices can obtain a position fix within a given accuracy
  • FFS: How often the UE may be required to obtain a position fix, which may be related to the required accuracy.
  • FFS: if position under this operation can be obtained by means other than GNSS that provides a comparable accuracy, e.g. pre-configuration for a fixed device. This may also include information other than positioning.
• GNSS-degraded: Refers to the network mode of operation which relies on devices being equipped with a GNSS receiver, the devices were able to obtain a position fix at some point in time, but the devices may not currently have a position fix within a given accuracy.
  • NOTE: The UE may be able to use the position fix for physical layer operation
• GNSS-free/GNSS-less: Refers to the network mode of operation which does not rely on devices being equipped with a GNSS receiver, or devices are equipped with a GNSS receiver but do not have a current position fix that can be used for physical layer operation.

6GR NTN targets to support GNSS-based operation, GNSS-degraded operation and GNSS-less/GNSS-free operation.

6GR NTN uplink time-frequency synchronization follows the same principle as NR NTN as baseline:

• The concept of "uplink synchronization reference point" is introduced in 6GR NTN.
• 6GR NTN provides satellite assistance information
• At least for GNSS-based operation, it is supported that UE uses its own location information + satellite assistance information to perform time-frequency pre-compensation.

Conclusion: NTN specific requirements and design for GNSS based operation (10.7.1)

As a general principle for 6GR NTN study:

• Under NTN agenda item, we will identify issues / requirements specific to NTN.
• Potential solutions to these issues / requirements may be studied under the NTN agenda
  • The outcome of this study may be discussed under other agenda items if common design is possible.
• These solutions may end up resulting in an extension of the TN design.
  • This may depend on the solution/issue/ requirement
• NTN specific solutions may be introduced when a common / extended design cannot meet the NTN requirements.
• When targeting a common design TN performance is prioritized.

6GR supports large scheduling offsets to accommodate the RTT introduced by the satellite channel. Further discuss how to realize these scheduling offsets:

• Option 1: Reuse the k_offset concept from NR as a baseline with potential modifications.
  • NOTE: Under this option, additional scheduling offsets (e.g. similar to K1/K2 in NR) may be supported for scheduling flexibility, which apply in addition to k_offset.
• Option 2: The large scheduling delays to accommodate the RTT are incorporated into the scheduling offsets (e.g. similar to K1/K2 in NR), which may or may not be common for TN and NTN

Conclusion: NTN specific requirements and design for GNSS based operation (10.7.1)

When reporting inputs for the link budget template and evaluation assumptions to RAN1#124b, companies are encouraged to provide them in an xls attached to their contribution following the format of the xls attached to R1-2601471.

Study NTN specific issues/requirements at least for the following aspects for 6GR NTN under this agenda:

• Uplink time-frequency synchronization
• HARQ issues: e.g. enable/disable HARQ feedback (which may include how to efficiently operate with HARQ feedback enabled/disabled), number of HARQ processes, etc.
• Timing relationships with large RTT.
• Coverage target
• Physical layer aspects of multi-satellite operation (including multi-orbit)
• Aspects related to Multiple beams per satellite
• Aspects related to satellite having more beam footprints than simultaneously active beams
• Other aspects are not precluded.

For NTN link budget template, RAN1 to take the TN link budget template as baseline with specific rows / values (including adding new rows) to be further discussed.

RAN1 will define evaluation parameters for at least the following combinations of satellite orbit and bands:

• S-band:
  • LEO 300, LEO 600, GEO
• Ka band:
  • [LEO 300], LEO 600, [LEO 1200], GEO
• Ku band:
  • LEO 1200, GEO NOTE 1: The evaluations for S band are expected to be similar to L-band. NOTE 2: This is only for the purpose of evaluations.

For 6GR AI/ML use cases identification/categorization, for each (sub-)use case proposed, proponent companies are encouraged to study and report the following:

• Definition of each (sub-)use case, including at least AI/ML model input/output
• The evaluation assumption, methodology, KPIs, benchmark, and preliminary simulation results
• Assumption on training types, e.g.,
  • offline training, online training/finetuning
  • Label construction (if applicable), including whether/how to obtain label data for model training
• Assumption on model location for inference, e.g., UE-sided model, NW-sided model, and two-sided model
• Collaboration/interaction between UE and NW, e.g.,
  • no collaboration/interaction
  • UE/Network collaboration targeting at separate or joint ML operation
• High level potential specification impact

SITE OWNER NOTE: This proposal was very long. For details of the tables, please, check the FL summary.

Observation For 6GR AI/ML use cases identification/categorization, [24 sources] provided preliminary simulation results and analysis on low overhead CSI-RS or CSI prediction with AI/ML.

• [23 sources] provided preliminary simulation results and analysis on frequency and/or spatial domain CSI prediction with sparse/low overhead CSI-RS with AI/ML.
• [6 sources] provided preliminary simulation results and analysis on CSI time domain prediction with AI/ML.
• [4 sources] provided preliminary simulation results and analysis on CSI prediction cross carrier/band/frequency block with AI/ML.
• [2 sources] provided preliminary simulation results and analysis on CSI prediction across analog beams with AI/ML.
• [one source] provided preliminary simulation results and analysis on Tokenized CSI prediction with linear projection as pre-processing.

Observation For 6GR AI/ML use cases identification/categorization, [23 sources] provided preliminary simulation results and analysis on low overhead DMRS with AI/ML receiver.

• [22 sources] provided preliminary simulation results and analysis on sparse orthogonal DMRS in frequency and/or time domain with AI/ML receiver.
• [11 sources] provided preliminary simulation results and analysis on superimposed pilot with AI/ML receiver.
• [5 sources] provided preliminary simulation results and analysis on DMRS free with AI/ML receiver.

Observation For 6GR AI/ML use cases identification/categorization, [13 sources] provided preliminary simulation results and analysis on CSI compression and feedback.

Observation For 6GR AI/ML use cases identification/categorization, [4 sources] provided preliminary simulation results and analysis on low overhead SRS with AI/ML. [1 source] provided preliminary simulation results and initial analysis on low PAPR SRS sequence design with help of AI/ML.

Observation For 6GR AI/ML use cases identification/categorization, [3 sources] provided preliminary simulation results and analysis on AI-enabled UL precoder indication.

Observation For 6GR AI/ML use cases identification/categorization, [3 sources] provided preliminary simulation results and analysis on AI/ML based waveform for PAPR reduction.

Observation For 6GR AI/ML use cases identification/categorization, [one source] provided preliminary simulation results and analysis on pathloss prediction in the spatial, temporal, and/or frequency domain, to use the predicted pathloss in UL (PUSCH/PUCCH/PRACH/SRS) power control.

• [one source] provided preliminary simulation results and analysis on UL closed-loop power control with an NW sided AI/ML model, where the model predicts the optimal power adjustment (or TPC command index) for the UE.
• [one source] provided preliminary simulation results and analysis on prior-information-aided DCI decoding.
• [one source] provided preliminary simulation results and analysis on lossless DCI compression.
• [one source] provided preliminary simulation results and analysis on early contention resolution in RACH.
• [one source] provided preliminary simulation results and analysis on sensing based RAN digital twin construction with NW-side AI/ML model.
• [one source] provided preliminary simulation results and analysis on AI/ML-enabled RAN digital twin with distributed model.

For study/evaluation of the performance and feasibility of AI/ML use cases in 6GR, at least the following may be considered

• Intermediate performance KPIs (e.g., SGCS), link level KPIs (e.g., BLER) and system level KPIs (e.g., throughput vs overhead), etc
• Computation complexity/latency (inference/monitoring)
• Power consumption, if feasible to evaluate
• Model size
• Data collection impact
• Scalability (refer to the examples in TR 38.843)
• Generalization performance
  • FFS on whether and how to consider realistic deployment scenarios
• Overhead/complexity associated with data collection, inferencing, performance monitoring, online/site specific fine-tuning, inter-vendor collaboration (if applicable)
• Online training/fine-tuning training latency, if feasible to evaluate
• Inter-vendor collaboration impact, if applicable Note: Details to be discussed per use case. Note: above aspects may be considered for both AI/ML and non-AI counter part

Endorse observation 2.1 ~observation 2.17 in R1-2508811 section 6. Note: this is to replace the corresponding observation in RAN 1 #122bis

From RAN 1 perspective, the following use cases can be matched to the identified primary agendas of RAN1

10.9 AI/ML for RAN#111 LS R1-2601598 is endorsed.