3GPP TSG RAN WG1 Meeting #120bis	R1-2501706
Wuhan, China, April 7 – April 11, 2025
Agenda Item:	9.4.1
Source:	Futurewei
Title:	Discussion on modulation aspects for A-IoT physical channel
Document for:	Discussion and Decision 

Conclusion
Proposal 1: CP handling method M2-2 Candidate 1 is not pursued any further. 
Proposal 2: For CP handling in R2D consider: 
Adopt Method Alt M1-1 for M=K: circular shift of the M chips before CP insertion at the reader side in addition to bit dropping at the device side.
FFS: K=8.
Proposal 3: For an OFDM-based waveform with subcarrier spacing of 15 kHz, Btx,R2D is the integer multiples of 180 kHz.
Proposal 4: Consider  chips with equal durations  within the OFDM symbol, not including the CP duration, i.e.,  .

Observation 1: The CP handing methods need to support large M values such as 16, 24 and 32 to achieve a comparable or better R2D data rate compared to UHF RFID.
Observation 2: The R2D Timing-Acquisition preamble waveform provides a reference point to Ambient IoT devices to detect the beginning of the NR OFDM symbol. The device uses this reference point to locate and discard the samples corresponding to the CP section.

Observation 3: Due to SFO in the device, the location of the reference point may drift as the device receives the R2D transmission. As a result, the device may discard samples not corresponding to the assumed CP section and degrade the detection performance 
Observation 4: Method Alt M1-2 is simple to implement and does not need a reference point as the Alt M1-1.
Observation 5: Method Alt M1-2 does not work for large M values when the CP length is not distinguishable from the length of OOK chip.
Observation 6: Method Alt M1-2 will result in non-constant chip length that may degrade the decoding performance.
Observation 7: Method Alt M2-1-2 may result in the duration of OOK chips at the beginning or end of an OFDM symbol different from other OOK chips and impact the device decoding performance.   
Observation 8: Method Alt M2-1-2 only works for cases where the OOK chip duration is longer than CP length. In another words, Method Alt M2-1-2 does not work for large M (>14) values for 15 kHz SCS.
Observation 9: The length of combining all CP sections from consecutive 14 OFDM symbols is exactly the length of an OFDM symbol. 
Observation 10: Compared to Method M1-1, with ideal timing Method M 2-1-1 Candidate 1 provides the detection performance improvement about 1dB for M=8 and M=16, about 2.5 dB for M=24, and over 15 dB for M=32.
Observation 11: The improvement that Method M 2-1-1 Candidate 1 provides increases with the M values. 
Observation 12: For M=24, with 1 sample time error Method M 2-1-1 Candidate 1’s improvement over Method M1-1 increases from ~2.5 dB to over 5 dB.
Observation 13: With preamble detection Method M 2-1-1 Candidate 1’s improvement over Method M1-1 is about 3 dB for M=8, 5 dB for M=16 and over 15 dB for M=32.
Observation 14: M2-2 Candidate 1 does not provide a solution to CP insertion in a standard OFDM signal generation process. It produces OFDM subcarriers not orthogonal to 5GNR subcarriers. 
Observation 15: When coexistence between NR and AIoT is needed in the future, sharing signal generation process in a base station between M2-2 Candidate 1 and regular NR waveform is very difficult if not impossible. 
Observation 16: Method M 2-1-1 Candidate 1 generates perfect OOK pulse string, yet the subcarrier orthogonality is kept.
Observation 17: Method M 2-1-1 Candidate 1 has relatively high overhead for small M, e.g., 31% for M=6. On the other hand, it has small overhead for large M, e.g. 5% for M=32 and 11% for M=16.

3GPP TSG-RAN WG1 Meeting #120bis	R1- 2501719
Wuhan, China, April 7th – April 11th, 2025

Agenda Item:	9.4.1
Source:	Ericsson
Title:	Modulation aspects for Ambient IoT
Document for:	Discussion, Decision
1	

Conclusion
In the previous sections we made the following observations:
Observation 1	It is feasible to generate OOK-4 using either CP-OFDM (with an optimization algorithm, e.g., LS) or DFT-S-OFDM (with DFT precoding) for any given M value.
Observation 2	For R2D CP handling Method 1, at least for Alt 1, the device needs to be aware of the boundary of the OFDM symbol (i.e., the beginning of the OFDM symbol) to determine the CP location.
Observation 3	Method Type 2 does not need to be aware of OFDM symbol boundaries, but this comes at the cost of spectrum inefficiency or the loss of subcarrier orthogonality.

Based on the discussion in the previous sections we propose the following:
Proposal 1	For OOK-4 generation, both CP-OFDM and DFT-S-OFDM are feasible for any value of M, and the choice between them is transparent to the device, via reader implementation choice.
Proposal 2	For OOK-4 generation,
Step 1: The time domain OOK signal is the M chips of one OFDM symbol.
Step 2: A chip is represented by L samples.
Step 3:  An N’-points DFT/LS is performed on the samples of one OFDM symbol to obtain the frequency domain signal.
Step 4: Map the frequency domain signal obtained by N’-points DFT/LS to the X subcarriers of Btx,R2D.
X is number of subcarriers allocated to A-IoT.
Step 5:  An N-points IDFT is performed to obtain the time domain signal.
Proposal 3	Support Method Type 1 Alt 1 and Method Type 1 Alt 2. The procedure for determining the boundary of an OFDM symbol using the R2D timing acquisition signal (R-TAS) needs to be specified.
Proposal 4	Method Type 2 Alt 1-1 is not supported (since it restricts scheduling and degrades spectral efficiency due to parity chips).
Proposal 5	Method Type 2 Alt 1-2 is not supported (since the non-uniform chip duration makes synchronization challenging)
Proposal 6	Method Type 2 Alt 2 is not supported (since it does not support subcarrier orthogonality, which prevents in-band deployment).
Proposal 7	M chips correspond to M ON/OFF segments within one A-IoT OFDM symbol before CP insertion.
Proposal 8	For R2D, in OOK-4 modulation for M-chip per OFDM symbol transmission, the maximum M should be 16.
Note: The number of allocated PRBs also imposes a limit on the maximum M.

3GPP TSG RAN WG1 #120b                                            R1-2501733                
Wuhan, China, April 7th – April 11st, 2025

Source:	TCL
Title:	Discussion on modulation aspects for Ambient IoT physical design
Agenda Item:	9.4.1
Document for:	Discussion and Decision

Conclusion
In this contribution, we provide our views on the feasibility and required functionalities of physical modulation for AIoT. The observations and proposals are listed as below:
Observation 1: For M=1, OOK-4 can be generated by CP-OFDM and DFT-OFDM, and for M>1, OOK-4 can be generated by DFT-OFDM.

Observation 2: The lowest threshold of ON bit for R2D signal transmission may impact on the performance of RF ED for energy detection method.
Observation 3: If edge method is used for start-indicator detection, the first level will not be detected by device but it can activate/wake-up device to receive following R2D signal, e.g., clock-acquisition.

Observation 4: For edge detection, if higher PAPR is generated in OOK-4 time domain waveform, device may detect wrongly OOK symbol.

Observation 5: Rel-19 do not need to consider the impact of multipath effect on frequency spectrum power.

Observation 6: Limited by device 1’s detection capability, before N’-points DFT performed, concentrated frequency spectrum could impact time domain ON/OFF chips with M*L samples , e.g., no flattened waveform or with higher PAPR, which is not suitable edge detection at device side.

Observation 7: Modulation like QAM/pi/2-BPSK/BPSK, scrambling of ON/OFF chips or sampling points, different sequence types or length, and predefined matrix or vector or sequence could be considered for handling concentrated frequency spectrum related problems.

Observation 8: N’=M*L=X represents without scrambling or redundant added or truncation/modification operation. N’=M*L*Q represents with scrambling or redundant added before DFT transferring and N’>X if with truncation/modification operation.

Observation 9: To ensure reader receives D2R signal correctly, reader could dynamically change mapping way based on SFO/channel/interference estimation to adapt OOK-4 signal center frequency to avoid channel fading.

Observation 10: 1/2 OFDM symbol duration or 1 OFDM duration as potential length has been discussed in 943.

Observation 11: If SIP length less than 1 OFDM duration and the start of SIP aligns with OFDM symbol n, padding filling is needed if CAP is also needed to align with OFDM symbol n+1. 

Observation 12: If SIP length less than 1 OFDM duration and the start of SIP aligns with OFDM symbol n, and there is no any padding filling and CAP is followed to SIP, which could not guarantee CP is inserted into the start of CAP. Because the last chip is OFF, device may not detect CAP part in OFDM symbol n if edge detecting used. 

Observation 13: If SIP length less than or equal to that of one OFDM symbol and the last of SIP aligns with OFDM symbol n+1, there is no need to fill padding and this case could ensure CP insertion normally.

Observation 14: If considering method 2-1-1 candidate 1, message may split into two chips at the start and end of one OFDM symbol after line coding.   

Observation 15: Small convex or concave duration easier happens after circular shift, which could be CP recognized failure. 

Observation 16: For different CAP pattern, there is different methods to calculate R2D chip duration 
For ON-OFF-ON-OFF pattern of CAP, chip duration equals to T1/2.
For ON-OFF-OFF-ON pattern of CAP, 
If OF-ON following ON-OFF-OFF-ON, i.e., ON-OFF-OFF-ON-OFF-ON, chip duration equals to T1/3.
If ON-OFF following ON-OFF-OFF-ON, i.e., ON-OFF-OFF-ON-ON-OFF, chip duration equals to T1/4.

Observation 17: One simpler method to calculate R2D chip duration is no need to regulate same direction of transition edge, then chip duration always equals to T1/2.

Observation 18: If CP inserted and transition edge occurs only at the start or only at the end of the CP, and subsequent chip with different level of CP, CP could be used for calculating chip duration and there is no special consideration for this case. 

Observation 19: If CP inserted and transition edge occurs during CP transmission, CP might be still used for calculating chip duration. Even CP has transition edge ON-OFF, the transition edge of end of CP (e.g., OFF-ON) could be used for calculating chip duration by combining subsequent transition edge of OFF-ON. 

Observation 20: For D2R transmission of AIoT, MCS-like is associated with D2R modulation (e.g., OOK and BPSK), channel coding (e.g., convolutional code), line code (e.g., with or w/o line code, repetition factor) or block/bit/chip-level repetition (e.g., repetition number). 



Proposal 1: Consider how to keep OOK-4 symbol with low PAPR or flat waveform during transmission.

Proposal 2: Each ON/OFF chip with randomized amplitude or phase after step 1 could be considered for handling concentrated frequency spectrum related problems.

Proposal 3: Consider potential methods for randomizing amplitude or phase of ON/OFF chip, e.g., modulation like QAM/pi/2-BPSK/BPSK, scrambling of ON/OFF chips or sampling points, different sequence types or length, and predefined matrix or vector or sequence, etc.

Proposal 4: How to solve signal distortion because of truncation operation for DFT-s-OFDM operation should be considered in WI phase.

Proposal 5: Consider combination of scrambling and adding redundant sequence to resist on the impact of truncation. 

Proposal 6: Discussion on the regulation of relationship between M, L, N’, X and N based on scrambling or redundant added before DFT or truncation/modification operation before IFFT.

Proposal 7: Select M value to ensure OOK-1/4 signal generation with flat waveform or lower PAPR.

Proposal 8: Consider how to map the out of DFT to subcarrier before IFFT.

Proposal 9: It is preferred that padding filling is up to implementation at reader side.

Proposal 10: If SIP length less than one OFDM symbol duration, the start or end of SIP needs to align with OFDM symbol. If the start of SIP aligns with OFDM symbol, padding filling is needed to ensure the start of CAP aligns with next OFDM symbol.

Proposal 11: If SIP length equals to one OFDM symbol duration, the start and end of SIP needs to align with OFDM symbol.

Proposal 12: Clarify whether CP length could be larger than transition edge period if transition edge used for detection at device side. 

Proposal 13: Support CP with ON level and M value chosen from {2,4,6,8} for method type 1-1 and 1-2. 

Proposal 14: If supported Method 2-1-1 candidate 1, the start of circular shift and redundant part added to avoid small convex or concave duration should be further considered.

Proposal 15: M2-2 should not supported for WI discussion.

Proposal 16: For CAP transmission,
Support CAP length is greater than 1 OFDM symbol, e.g., 2 OFDM symbol duration
Support at least first OFDM symbol of CAP to insert CP
If CAP occupy multiple OFDM symbol, consider how/whether to design CP insertion for subsequent OFDM symbol.

Proposal 17: Support following methods for R2D chip calculation,
Option 1:  , where C represents a R2D chip duration
From device perspective, whether to define the following on top of above options
 
Where,  is a floating chip duration and FFS: 

Proposal 18: When CP inserted and transition edge occurs during CP transmission, CP insertion will impact the chip duration calculation. 

Proposal 19: Support CP inserted used for chip duration calculation if the level of CP is different from that of following pattern in CAP.
Proposal 20: Discuss the definition of MCS-like for D2R transmission after SFS discussion of 942. 

3GPP TSG RAN WG1 meeting #120bis                                   R1-2501760
Wuhan, China, April 7th - 11th, 2025
 
Source: 	ZTE Corporation, Sanechips
Title:	Discussion on Ambient IoT modulation
Agenda Item:	9.4.1
Document for:	Discussion and decision

Conclusion
In this contribution, we discuss the Ambient IoT physical layer design and have the following observations and proposals.
Observation 1:It will result in an inferior CAP detection performance because of the close and overlapped range of chip duration between adjacent M values considering SFO under different device sampling rates.
Observation 2:CP handling method type 1 (i.e., removal of CP at device) can mitigate the potential impacts of non-constant chip duration, but without any other enhancement, it will cause performance loss in the cases of inaccurate timing synchronization, timing error accumulation caused by false determination of the transition edge, especially for larger M values.
Observation 3:Alt M2-1-1 candidate 3 is applicable to large M value.
Observation 4:For Candidate 5, it is assumed the R2D data transmission starts from the last chip in the first OFDM symbol and has no additional overhead, i.e., the first odd number of chip(s) in the first OFDM symbol can be reserved for preamble transmission. 
Observation 5:For Candidate 5,, in the case of M≤16 that CP length is less than or approximately equal to a chip duration, the CP concatenated with either the first or the second chip of the Manchester codeword doesn’t introduce false transition edge around CP. 
Observation 6:For candidate 5 with extended last chip duration, resource efficiency does not reduced because M value is kept unchanged and there is no redundant chip among M chips.
Observation 7:Because there are no false transition edge introduced by CP and both the last chip and CP are similar to two chip duration for M=24, a update of the threshold for the edge detection around the CP is enough and CP removal operation (i.e. Alt M1-1 and Alt M1-2) is not required, e.g. interval between two valid transition edges is either in the range of 4*T~5.5*T or in the range of 1.5*T~2.5*T for the detection of all codewords.
Observation 8:Under the case of TBS=20bits, Candidate 5 has better BLER performance than Alt M1-1, i.e. almost 1dB and 2dB SNR gaps for M=16 and M=24 respectively. 
Observation 9:Under the case of TBS=96bits, 1) Candidate 5 has almost 0.5dB performance gains compared to Alt M1-1 for M=16; 2) for M=24, Candidate 5 achieves BLER performance of SNR=20dB@10% BLER, while Alt M1-1 has an error floor.
Observation 10:For CP handling method Candidate 5 with extended last chip duration, there are one last chip with 10 samples, mod(128-10, 24-1) = 3 chips with  samples and 20 chips with  samples in an OFDM symbol if IDFT size N equals to 128 for M=24.

Proposal 1:The maximum M value should be 24 .
Proposal 2:For R2D, in order to avoid the ambiguity issue, it is proposed that M values of 8 and 16 should be removed.
Proposal 3:For R2D, specify the minimum Btx,R2D # of PRBs associated to each selected M value in TR 38.769.
Proposal 4:It is necessary to ensure that no false transition edge is introduced by CP.
Proposal 5:In order to avoid the false transition edge during a CP for M>=24, the last chip duration in an OFDM symbol can be extended to be the same as the CP duration (e.g. 10 samples), and the chip duration of some other chips except for the last chip in an OFDM symbol is reduced from 6 samples to 5 samples.
Proposal 6:Candidate 5 is proposed to be used for CP handling because of the advantages that it does not introduce additional overhead and provides better BLER performance. 
Proposal 7:For M>=24, Candidate 5 with extended last chip duration is proposed to provide a better performance by avoiding the false transition edge during a CP.
Proposal 8:Specify the following procedure for R2D waveform generation,without restrict on the value of L, N’, X and N.
1. The time domain OOK signal is the M chips of one OFDM symbol.
2.  A chip is represented (e.g. upsampled) by L samples
3.  An N’-points DFT is performed on the samples of one OFDM symbol to obtain the frequency domain signal.
4. Map the frequency domain signal obtained by N’-points DFT to the X subcarriers of Btx,R2D. 
5. An N-points IDFT is performed to obtain the time domain signal.
Proposal 9:Chip duration should be clarified because it needs the samples of each chip under the following cases where device needs to know at least the samples within last chip, especially for the case that IDFT size (N=2^n) can not be divided by M values or the CP handling method candidate 5 is adopted.
Proposal 10:Define the R2D chips within an OFDM symbol as below: 
if CP handling method Candidate 5 with extended last chip duration is used for M>=24, the last chip duration = CP duration with  samples and other chips' duration =  or ;
otherwise, chip duration =  or , wherein .

3GPP TSG RAN WG1 #120b	R1-2501776
Wuhan, China, April 7th - 11th, 2025

Source:	Nokia
Title:	AIoT Physical channels design - modulation aspects
Agenda item:	9.4.1
Document for:	Discussion and Decision

Conclusion
In this contribution we have made the following observations and proposals:
Observation 1:	The AIoT device can differentiate the detected chips as a chip used for data communication (i.e., On-chip or Off-chip) and a padding chip.
Observation 2:	The unknown information of modulation scheme affects receiving operation at the reader side. It is unclear how often the reader should attempt to detect which modulation scheme is used for a given AIoT device.
Observation 3:	The receiving operation at the reader differs depending on whether the modulation scheme indication feature is supported.

Proposal 1: RAN1 should decide the maximum value of M solely based on the required maximum data rate. E.g., M=24 would provide comparable performance to RFID.
Proposal 2: RAN1 supports a subset of M values listed in Table 6.1.1.4-1 of TR 38.769. E.g., M= 2, 6, 12, and 24.


Proposal 4: Support reader-side indication of the modulation scheme - either OOK and BPSK – to the AIoT devices. 

3GPP TSG RAN WG1 #120 bis	R1-2501806
Wuhan, China, April 7th – 11st, 2025

Source: vivo
Title: Discussion on Modulation Aspects of Physical Channels Design 
Agenda Item: 9.4.1
Document for: Discussion and Decision

Conclusion
In this contribution, we discuss the remaining issue for modulation aspects of physical channels design. And we have the following observations and proposals.
Observation 1: M=16 is sufficient to achieve comparable data rate compared with that in RFID from reader to AIoT device link.
Observation 2: M=32 suffers material performance degradation, with the presence of CP, BLER error floor occurs even at 20% BLER. 
Observation 3: Although M=1 R2D transmission uses 1/2 half data rate compared with M=2 transmission, the average transmission power for R2D with M=2 is twice of R2D with M=1, which means M=1 cannot achieve coverage gain even with lowest data rate.
Observation 4: When M=12, ratio of ON/OFF chip in PRDCH could be very close to 3:1, which lead to high FAR for SIP detection.
Observation 5: M = 6 and M=12 for CAP and data part would introduce new chip length for D2R, in addition to D2R chip length determined by scaling of R2D chip length corresponding to M = {2,4,8,16}.
Observation 6: Method Type 1 works effectively for all M values, and it is up to device implementation on Alt M1-1 and/or Alt M1-2 is/are used. There are no RAN1 specification impacts. 
Observation 7: Method Type 2 Alt M2-1-1 Candidate 1 is applicable to all M values, but, :
Higher device implementation complexity: different CP handling for different M values
Higher gNB implementation complexity: different CP handling for different M values, complicated determination of number of samples for cyclic shift and parity bits, which varies with OFDM symbol index, and change of implementation of Tx hardware after IFFT output, due to additional cyclic shift.
Degraded detection performance due to imbalance power across OFDM symbols
Lower spectrum efficiency due to parity chips
Observation 8: The actual M value for Method Type 2 Alt M2-1-1 Candidate 2 is not clear and large, and requires more RBs for R2D Tx BW.
Higher device implementation complexity: different known signal for different M values to be detected.  
Higher gNB implementation complexity: different known signal generation for different M values. 
Lower spectrum efficiency due to parity chips in time domain and larger number of PRBs to generate much shorter ON/OFF duration for the known signal. 
Detection performance highly depends on the design of the known signal but no detailed design for the known signal is available yet.  
Observation 9: Method Type 2 Alt M2-1-1 Candidate 3 can be workable for all M values (except M=2, if parity bits to be added at both start and end of OFDM symbol), but: 
Degraded detection performance due to imbalance power across OFDM symbols, and wrongly determination of end of R2D due to confusion between irregular chip duration by parity chips and postamble for R2D (if supported).
Lower spectrum efficiency due to parity chips. 
Observation 10: Method Type 2 Alt M2-1-1 Candidate 4 using alerting M to M+1 in adjacent OFDM symbols is only workable for M=4 case, Method type 1 is needed for other M values: 
Higher device implementation complexity: different CP handling for different M values
Higher gNB implementation complexity: different CP handling for different M values, determination of M or M+1 among OFDM symbols based on information bit value. Some M+1 value may not feasible to be generated considering limitation on pre-DFT size.
M+1/M-1 is not candidate value for R2D transmission.
Degraded detection performance due to imbalance power across OFDM symbols
Inaccurate clock calibration due to variable OOK chip length
Observation 11: Method Type 2 Alt M2-1-1 Candidate 5 is workable for M < 6, while Method type-1 is needed for other M values.
Higher device implementation complexity: combination of two different Method types for the same M value
Higher gNB implementation complexity: complicated power allocation across OFDM symbols
Degraded detection performance due to imbalance power across OFDM symbols, ON chip power is reduced compared with balanced number of ON/OFF chips within a OFDM symbol.
Inaccurate clock calibration due to variable OOK chip length.
Observation 12: Method type 2 Alt M2-2 can be workable for all M values, but post IFFT module should be changed compared with existing OFDM generator, introducing additional complexity for gNB implementation.
Observation 13: For Method type 2-1 candidate 1/4/5, Method type 1 is still needed since the candidate signal design is applicable for subset of the M values, which also means Method type 1 is feasible. 
It is not clear why other candidate scheme is still needed if Method type 1 can work well.


Proposal 1: M=32 is not supported for Rel-19 AIoT.
Proposal 2: The maximum M value for R2D is 16.
Proposal 3: M=1 for R2D is not supported.
Proposal 4: For reliable SIP detection, M=12 is excluded to avoid ON/OFF or OFF/ON pattern with length ratio close to 3:1.
 Proposal 5: M=6 should be used for SIP part design, considering following design criterion.
One OFDM symbol contains multiple ON/OFF
The multiple ON/OFF pattern includes 3:1 ON/OFF or OFF/ON ratio
M value can be supported by R2D transmission BW of 1 RB
Proposal 6: M=6 and M=12 should not be supported for CAP part and R2D data part.
Proposal 7: M = {2, 4, 6, 8, 16} is supported for R2D transmission, and
M=6 is only used for SIP part
M = {2, 4, 8, 16} is used for CAP part and data part.
Proposal 8: CP Handling method type 1 is sufficient. Other CP handling candidate schemes are not supported.
Proposal 9: R2D chip length is defined as .
Proposal 10: The content of padding is up to reader implementation and transparent to A-IoT device.
From R2D Tx perspective, up to reader implementation to fill in the remaining chips to align with OFDM symbol boundary.  
From Rx perspective, device can identify the end of R2D transmission per RAN2 agreements by the MAC header.
Proposal 11: RAN1 does not specify details of R2D waveform generation except for 
The minimum BW for R2D transmission, and 
Time domain structure of CP + M-chip per OFDM symbol duration for R2D transmission.
DFT-S-OFDM baseband generation formula maybe needed in spec, with some parameter e.g., a(k), left un-defined and up to implementation. 

3GPP TSG RAN WG1 #120b                                                                R1-2501868
Wuhan, China, April 7th – 11th, 2025

Agenda Item:	9.4.1
Source:	Spreadtrum, UNISOC
Title:	Discussion on modulation aspects of physical channels design for Ambient IoT
Document for:	Discussion and decision

Summary
In the contribution, we provides our view on the ambient IoT device architectures, and propose that,
Proposal 1: When the generated number of chips for the R2D transmission does not fully occupy the last OFDM symbol, padding is used to align with the boundary of the last NR OFDM symbol for in-band operation. The content of padding is up to reader implementation and transparent to A-IoT device.
Proposal 2: The text in NR specification for OFDM baseband signal generation and transform precoding can be the baseline for capturing DFT-s-OFDM signal in Ambient IoT specification.
Proposal 3: The step for DFT-s-OFDM signal generation should not be captured in the specification (TS), and the value of parameters L(number of upsampling points), N (number of DFT points), and N’ (number of IDFT points) can up to reader implementation .
Proposal 4: M values with 1, 4, 16, 24 are supported.
Observation 1: For small M values, e.g. 4, the impacts of CP for device demodulation is minor or can be negligible.
Proposal 5: When M value is small (e.g., M = 4), as CP length is much smaller than a chip length, CP handling is not needed or can be up to device implementation.
Proposal 6: The CP handling method which does not retain subcarrier orthogonality (i.e., Alt M2-2) should not be further considered.
Proposal 7: Method Type 1 (Alt M1-1 and/or M1-2) can be supported by device implementation with no specification impact.
Proposal 8: One solution from Alt M2-1 should be supported for CP handling for large M values (e.g. 16, 24).
Proposal 9: Alt M2-1-1 (candidate 3) should be supported when M is 16 and/or 24.

3GPP TSG RAN WG1 #120bis			R1-2501992
Wuhan, China, April 7th – 11th, 2025
  
Source:             CATT
Title:                  Ambient IoT physical channel design and modulation
Agenda Item:    9.4.1
Document for:  Discussion and Decision	

Conclusion
In this contribution, the general aspects of physical layer design for NR Ambient IoT communication system are discussed. We have the following observations and proposals:
Proposal 1: Since the A-IoT D2R does not have the orthogonality with UL NR signal, NR A-IoT R2D signal would not need to maintain a strict orthogonality with legacy DL NR signal.
Proposal 2: Method Type 2-M2-2, which extends the legacy OFDM symbol duration without NR CP operation, should be supported for R2D signal transmission due to its lower complexity and more efficient spectrum resource utilization compared to other methods.
Observation 1: For Method Type 2-M2-2, existing mechanisms can resolve or mitigate the inter-system interference between the A-IoT system and the NR system as follows:
Solution 1: The NR DL/UL signal and the A-IoT R2D/D2R signal can be transmitted using TDM.
Solution 2: The legacy NR UE and A-IoT devices can be deployed in different deployment scenarios, such as deploying the A-IoT reader/devices indoor while the NR UEs outdoor.
Solution 3: Using large guard band to mitigate the interference between A-IoT system and NR system.
Solution 4: gNB could utilize some existing mechanisms, such as using the pulse shaping filter, to minimize the interference from the R2D signal/channel to the adjacent NR channels during the transmission.
Proposal 3: The Method 1-1 and 1-2 of Type 1 CP handling, i.e. Removal of CP at device without explicitly specification of transmitter to simplify the A-IoT device detection should not be supported in Rel-19.
Observation 2: For Method Type 2-Alt M2-1-2, the overheads caused by the parity-check chip(s) (i.e. last chip(s)) and CP are 29.9% and 53.3% for M = 8 and 4, respectively, when the CP is not included in first chip.
Proposal 4: Method Type 2-Alt M2-1-2, where the first OOK chip(s) and the last OOK chip(s) in an OFDM symbol are the same, should not be supported due to the unnecessary signal overhead it introduces.
Observation 3: For the subcarrier orthogonal candidate method, the new transition edge of CP would be introduced for M larger than 16, which will degrade the synchronization performance and increase the A-IoT device complexity.
Proposal 5: The comparison and analysis for the candidates of CP handling in following table should be considered for the CP handling method determination.
Table: The comparison and analysis for the candidates of CP handlings

Proposal 6: If the orthogonal method is adopted, the Candidate 6 in Method Type 2 (Alt M2-1-1) could be supported, i.e. the end of an OFDM symbol corresponding to the CP length will be set to high voltage transmission for M equals to 16.
Proposal 7: The R2D chip duration should be defined as that of an OOK chip of OOK-4 signal, which is equal to 1/M of an OFDM symbol length (excluding the CP), where M≥1.
Proposal 8: For Device 1, the R2D signal should only support the following M values: {2, 4, 8, 16}.
Proposal 9: The durations related to the D2R transmission should be clarified as following:
D2R information bit duration: It is defined as the D2R information bit for the D2R transmission before the channel coding.
D2R bit duration: It is defined as the duration of each coded bit after FEC, which is independent from the small frequency shift (SFS) factor.
D2R chip duration: It is defined as the actual transmission duration for the D2R transmission after the SFS, which has the relationship with the D2R bit duration that D2R bit length = 2 * R * chip length. 
Proposal 10: For Device 1, the bit rate of PDRCH without SFS consideration should not be higher than 240kbps with the equivalent M value of 16.
Proposal 11: The bit duration would need to inform the device via the implicit or explicit indication in general, although it is critical for the Msg1 in CBRA procedure for FDMA.
Proposal 12: The three options for D2R transmission indication based on the certain chip duration indication should be specified as following:
Option 1：The D2R chip duration and corresponding SFS factor R values set are predefined. The two-stage D2R transmission indication is carried in PRDCH control information. The D2R bit duration could be obtained via the indicated chip duration and the minimum R value in the R value set. 
Option 2: The D2R chip duration and the SFS factor R are carried in PRDCH control information. The D2R bit duration could be obtained via the indicated chip duration and the minimum R value according to the predefined R value set or TBS related information.
Option 3:  The D2R chip duration is implicitly indicated by deriving from the control signaling of coded bit duration and SFS factor R.
Observation 4：For D2R backscattering from Device 1, the TBCC with the BPSK modulation could obtain about 7.1dB performance gain over the TBCC with the OOK modulation for BLER at 10-2.
Observation 5: The D2R receiver can detect OOK and BPSK using the unique processing procedure, i.e. BPSK demodulation procedure, which cannot degrade the BLER performance of OOK modulation and BPSK modulation without phase discontinuity issue.
Observation 6: The phase discontinuity issue of OOK demodulation by the reader using the BPSK demodulation method after DC removal would severely degrade the OOK demodulation performance if the reader does not know the D2R waveform is OOK or BPSK.  
Proposal 13: In order not to degrade the OOK demodulation performance, the reader should be aware of the modulation scheme of OOK or BPSK for the D2R transmission.
Proposal 14: For modulation quality, a less stringent EVM requirement should be defined for Device 1 than that in NR, which can be determined in RAN4. 
Proposal 15: The TBS tables should be introduced for simple indication of TBS and the corresponding TTI, which are aligned with the NR time domain resource assignment.
Proposal 16: Adopt the following TBS tables for PRDCH and PDRCH: 
The TBS table for PRDCH (3 bits TBS indication)

The TBS table for PDRCH (3 bits TBS indication)

Proposal 17: From the device perspective, the specific transmission bandwidth for R2D signal/channel, which is greater than or equal to the minimum Btx,R2D value, should be up to gNB implementation.
Proposal 18: The FDMA would not be supported or only support a limited number of SFS factors under the 5MHz system bandwidth, when the data rate is higher than 150kbps.
Proposal 19: The all “1” modulation should be specified for the DFT-s-OFDM based OOK-4 waveform, since the centralized power distribution at a few frequency domain samples is not sensitive for the subcarriers mapping.

3GPP TSG RAN WG1 #120-bis          						R1-2502016
China, Wuhan, April 7th – 11th, 2025

Source:	Tejas Networks Ltd.
Title:	Discussion on modulation aspects for A-IoT physical channel
Agenda item:	9.4.1
Document for:     Discussion and Decision

Conclusion
This work includes the modulation aspects of A-IoT Physical Channel, estimation of optimum M-value for R2D signal generation, and CP handling for R2D. We have made the following observations and proposals related to the above-mentioned aspects of A-IoT:

Observation 1: For M = 24, the data rate is above 100kbps and for M = 16, the data rate is below 100kbps.
Proposal 1: We propose M = 24 as the maximum M value, as it provides data rate > 100kbps.
Observation 2: As R2D signals are transmitted using OOK-4 generated by OFDM modulator, it is important to define the M-value for OOK-4 modulation for all DL message signals.
Proposal 2: we propose the following three options to decide the optimum M value. 
Option 1: Msg0 contains known preamble from which the device can estimate the channel condition and choose an optimum M value from a lookup table stored in the device and convey the same through msg.
 
Option 2: Msg1 contains known preamble from which the reader can estimate the channel condition and choose an optimum M value from a lookup table stored in the reader and the M value will be used for subsequent DL transmissions.

Option 3: An optimum value can be decided based on the following parameters:
Received signal strength of the UL message signals received form the device
Block error rate
Data rate and latency
Observation 3: In order to decode the PRDCH frame including the control information, the device should have the information of CP length of the OOK-4 modulated signal generated from OFDM modulator even before receiving the control information. 
Proposal 3: CP length can be derived as a function of M-value, which is detected from the length of clock-acquisition signal. The device estimates the M-value of the received PRDCH signal from the clock-acquisition signal and derive the CP length form the M-value.

3GPP TSG RAN WG1 #120bis			R1-2502022
Wuhan, China, April 7th – 11th, 2025

Agenda item:		9.4.1
Source:	China Telecom
Title:	Discussion on physical channels design about modulation aspects for Ambient IoT
Document for:		Discussion

Conclusion
In this contribution, we have the following proposals:
Proposal 1: For R2D transmission, support DFT-s-OFDM waveform generation for OOK-4 modulation.
1.	The time domain OOK signal is the M chips of one OFDM symbol.
2.	A chip is represented (e.g. upsampled) by L samples
L samples are generated from M-bits and up to reader implementation
3.	An N’-points DFT is performed on the samples of one OFDM symbol to obtain the frequency domain signal.
N’=128 or equal to X, as a starting point
4.	Map the frequency domain signal obtained by N’-points DFT to the X subcarriers of Btx,R2D. 
X subcarriers is related to the bandwidth Btx,R2D
5.	An N-points IDFT is performed to obtain the time domain signal.
N can be same or different as N’.
If N is not equal to N’, truncation or other additional modification is needed at reader side.

Proposal 2: For R2D transmission, no need to support M value larger than 24 for A-IoT device 1.
Proposal 3: Update the table of M value and the associated minimum Btx,R2D value as follows.

Proposal 4: For CP handling of A-IoT R2D transmission, only normal CP is supported.
Proposal 5: If any part of R2D preamble needs CP handling, the same CP handling solution applies for both R2D preamble and PRDCH.
Proposal 6: For CP handling if only one solution is allowed, support to remove CP at device without specified transmit-side.
Using Alt M1-1 and/or Alt M1-2 is up to device implementation as per TR 38.769
Proposal 7: For R2D CP handling, Alt M2-2 (as per TR 38.769) is not considered in Rel-19.
Proposal 8: From the transmitter perspective, the start of R2D transmission is assumed to be aligned with the boundary of an NR OFDM symbol (including the CP) for in-band/guard-band operation.
Proposal 9: For OFDM-based OOK-4 waveform generation, at least padding can be used to keep the end alignment in the last NR OFDM symbol.
The content of padding is up to reader implementation and transparent to A-IoT device.
Proposal 10: The NR symbol boundary alignment is not necessary for D2R transmission.
Proposal 11: Both the first and last symbol occupied by D2R transmission can be regarded as the allocated reception resource at reader side.

3GPP TSG RAN WG1 #120bis	R1-2502068
Wuhan, China, April 7th – 11th, 2025

Agenda Item:	9.4.1
Source:	NEC
Title:	Discussion on modulation aspects of ambient IoT
Document for:	Discussion and Decision
1	

Conclusion
In this contribution, we give our views on modulation aspects of ambient IoT. We propose that:
Observation 1: M = 32 could be used only when reader and device are close with each other where the delay spread of channel are expected much less than time duration of one chip.
Proposal 1: To support the required data rate of ambient IoT’s service, the maximum M value should be 32.
Proposal 2: For CP handling for R2D transmission, support Alt M2-1-1 candidate 3. The padding chip is OOK-ON chip.
Proposal 3: DFT-s-OFDM waveform generation by frequency domain repetition method is specified, the steps are:
1) The time domain OOK signal is the M chips of one OFDM symbol.
2) An M points DFT is performed on the samples of M chips to obtain M point the frequency domain signal.
3) Map the M point frequency domain signal obtained by M points DFT to the X subcarriers of Btx,R2D. 
M to N point mapping is block repetition, i.e., ak = f((k - X/2 mod M), k = 0, …, X-1, where a is X subcarriers and f is M point frequency domain signal obtained by M points DFT.
4) X point base sequence (which is implementation by company and not specified) is multiplied on X subcarriers
5) NR OFDM baseband signal generation of X subcarriers is applied.
Proposal 4: Different Tx/Rx bandwidth assumption has impact on RAN1’s design, e.g., whether FDD full duplex could be assumed, whether subcarrier orthogonality CP handling is needed. RAN1 should align the view on the bandwidth assumption.
Proposal 5: Padding can be inserted into R2D postamble to meet the last OFDM symbol boundary.

3GPP TSG RAN WG1 #120bis		R1-2502123
Wuhan, China, April 7th – 11th, 2025
Agenda Item:	9.4.1
Source:	Fujitsu
Title:	Modulation for R2D and D2R
Document for:	Discussion
. 

Conclusion
In this contribution, we discussed CP handling and the value of M of OOK modulation in R2D. We have the following proposals and observations:
Proposal 1: Do not support Alt M2-2 for the CP handling in Rel-19.
Proposal 2: Support Alt M1-2 and Alt M2-1-2 for the CP handling in Rel-19 unless any infeasibility is confirmed.
Proposal 3: The maximum M value for device 1 is 12 or 16.

Observation 1: 
The usage of the CP of AIoT R2D signal is mainly to protect the NR DL signal adjacent to the AIoT R2D signal.
The CP inserted into the AIoT R2D signal does not conduce to the detection of the R2D signal at AIoT device side, because
AIoT devices have no capability to utilize the orthogonality provided by the CP to remove the interference coming from the adjacent NR DL signal.
The orthogonality is in frequency domain rather than in time domain; and,
The AIoT devices cannot demodulate the received signal in the frequency domain, which obviously exceeds its capability. 
The AIoT devices are equipped with RF ED only and without any high-quality passband filter. 
The AIoT devices can neither extract the R2D signal from the mixed-up signal nor suppress the impact of the NR DL signal.
Observation 2: Another potential advantage of inserting CP into the AIoT R2D signal may be that the timeline for producing NR DL signal at the gNB side can somehow be reused for producing the AIoT R2D signal. This kind of reusage is beneficial to minimize the development workload for the needful update of the gNB to support the AIoT system.

3GPP TSG RAN WG1 #120bis	R1-2502160  
Wuhan, China, April 7th – 11th, 2025

Source: 	CMCC
Title:	Discussion on modulation aspects of physical channel design
Agenda item:	9.4.1
Document for:	Discussion/Decision

Conclusions 
In this contribution, we give our consideration on modulation aspects of A-IoT physical layer design, and the following proposals and observations are made:
Observation 1: Considering the offset of sampling clock, it is difficult for a device to determine the M value based on the sampling points per chip, especially if adjacent M values are supported.
Observation 2: The interpretation of "one solution for CP handling" in the WID is that either subcarrier orthogonal solution or non-orthogonal solution is supported, and for a given M value, supporting multiple solutions are not allowed.
Observation 3: Discussion on combination of different alternatives to solve the CP issues under different M values is in the scope. And it is more important to consider low and medium M value cases.
Observation 4: The Candidate 4 of Alt M2-1-2 with  can avoid false rising/falling introduced by CP:
Change  case to  case by changing OOK-4 M value to adjacent  for OFDM symbol n.
where a, b and c is the last chip of symbol n, first chip of symbol n+1 and last chip of symbol n+1. 
Observation 5: The following Method Type 2, Alt M2-1-2-Candidate 5 with  can avoid false rising/falling introduced by CP:
Odd number of chips of PRDCH is transmitted in the first OFDM symbols, while M value is kept unchanged during the remaining OFDM symbols.
Proposal 1: The minimum number of PRBs that are used for reader Tx (Minimum Btx,R2D # of PRB) is not specified in RAN1. It can be up to implementation. However, if Minimum Btx,R2D # of PRB is specified in RAN1, it is proposed as follows,
Table 2.1-1. M values and the associated minimum Btx,R2D value
Proposal 2: Down-selection of M values can be considered.
For low R2D data rate, M=2 can be supported.
For medium R2D data rate, M=6 or 8 can be supported.
For large R2D data rate, M=24 can be supported.
Proposal 3: The followings are considered in the further discussion, 
Two impedance states are enough for either OOK or BPSK modulation.
Phase continuity of two ON chips are assumed to be kept for OOK modulation.
Proposal 4: Do not support both orthogonal and non-orthogonal CP handling solution, down-selection from orthogonal and non-orthogonal CP handling solution is needed.
Proposal 5: For CP handling, Alt. M2-2 which cannot retain subcarrier non-orthogonality is precluded.
Proposal 6: For M<6, support Method Type 2, Alt M2-1-2 to solve CP issues, both Candidate 4 and Candidate 5 can be further studied for down-selection.
Proposal 7: For M≥6, Method Type 1 with Alt M1-1 can be supported.
Proposal 8: In R2D, a chip
Corresponds to one modulated symbol for OOK-4.
Chip duration = (1/M) × {OFDM symbol duration excluding CP part}.
Proposal 9: For R2D link, from device perspective, floating chip duration that  is valid, and FFS .
Proposal 10: From device point of view, the end of R2D transmission is not required to be aligned with the boundary of an NR symbol for in-band/guard-band operation. It is up to reader implementation.

3GPP TSG-RAN WG1 Meeting #120bis	R1-2502233
Wuhan, China, April 7-11, 2025

Agenda Item:	9.4.1
Source:	Huawei, HiSilicon
Title:	Physical channels design on modulation
Document for:	Discussion and Decision 

Conclusions
In this contribution, the physical channels design on modulation for Ambient IoT are discussed and following observations and proposals are made accordingly.
Observation 1: M = 32 with minimum transmission bandwidth 4 PRBs cannot achieve BLER 10% and cannot work well for the R2D transmission.
Observation 2:  The M value set of M = {2, [6 or 8], 24} provides almost evenly SNR gap for the R2D transmission.
Observation 3:  If the minimum PRB of M = 6 is updated to 2 PRBs, the performance and transmission efficiency for M = 6 and M = 8 is close to each other.
Observation 4: RAN1 should down-select between M = 6 and M = 8 for the M value set determination.
Observation 5:  If M values with small gap e.g., M = 6 and M = 4, M = 8 are all in the M value table, false detection rate performance of the clock acquisition will be degraded.
Observation 6: Similar as RFID, the clock for receiving R2D chips corresponding to any of M value is up to device implementation.
Observation 7: We consider to support the M value set of {2, 6 or 8, 24} in the following aspects:
Including M = 24 is necessary since it can achieve comparable chip rate as UHF RFID and good performance.
Even SNR gap with the M value set {2, 6 or 8, 24} is sufficient for different coverage/data rate.
No FDR performance loss for the clock acquisition performance with sufficient gap among M values in the set of {2, 6 or 8, 24}.
Observation 8: Alt M1-1 is achievable within the implementation complexity for device 1.
Observation 9: Alt M1-1 is based on the last rising or falling edge detected in the OFDM symbol.          
Observation 10: For Alt M1-1, CP length difference among OFDM symbols leads to only negligible performance loss even when M is as large as 24.
Observation 11: Alt M1-2 is not a complete solution to CP handling, since it does not work for M values which result in CP length > chip length.
Observation 12: Alt M2 does not require the device to identify the CP location and remove/skip the CP part.
Observation 13: For R2D CP handling, Alt M2-2 will destroy the subcarrier orthogonality.
Observation 14: For R2D CP handling, if Alt M2-2 is used for standalone operation, it certainly adds complexity to the system and may increase device implementation complexity to distinguish such difference or impact on the R2D receiving if such difference is agnostic to device.
Observation 15: Alt M2-1-1 Candidate 3 only and Alt M2-1-2 Candidate 5 only suffer performance loss ~3 dB and ~2 dB compared with Alt M1-1 only for small M value, e.g., M = 6.  
Observation 16: The combined methods of Alt M2-1-1 Candidate 3 and Alt M2-1-2 Candidate 5 with Alt M1 do not provide additional performance gain but slight performance loss ~0.8dB compared with Alt M1 only for small M value, e.g., M = 6. 
Observation 17: For R2D transmission, Alt M1 only yields the best performance for small M value, e.g., M = 6.
Observation 18: The combined methods of Alt M2-1-1 Candidate 3 with Alt M1 and the combination method of Alt M2-1-2 Candidate 5 with Alt M1 has very close performance compared with Alt M1 only for the medium M value, e.g., M = 12. 
Observation 19: For R2D transmission, Alt M2-1-1 Candidate 3 combined with Alt M1 yields ~1.2 dB pure performance gain and ~0.82 dB performance gain considering transmission efficiency loss compared with Alt M1 only for large M value, e.g., M = 24.
Observation 20: For R2D transmission, Alt M2-1-2 Candidate 5 combined with Alt M1 yields close performance and slight performance gain under BLER 10-1 compared with Alt M1 only for large M value, e.g., M = 24.
Observation 21: Padding is needed after the end of R2D transmission for the alignment of OFDM symbol boundary.
Proposal 1: The maximum M value is required to ensure the corresponding R2D chip length comparable with UHF RFID minimum chip length.
Proposal 2: M = 32 is not supported in the M value set for the R2D transmission.
Proposal 3: The maximum M value is 24 for the R2D transmission.
Proposal 4: Specify the following {M, minimum Btx,R2D} values if the specific M value is supported for R2D:
Proposal 5: The M value set of {2, 6 or 8, 24} is sufficient for the R2D transmission.
Proposal 6: For R2D CP handling, Alt M1-1 is supported for all M values.
Proposal 7: For R2D CP handling, Alt M1-2 is a part of a likely implementation of Alt M1-1 and can be used for small M values.
Proposal 8: For R2D CP handling, Alt M2-2 is not supported.
Proposal 9: For R2D CP handling, Alt M1 is supported for large M value, e.g., M = 24 since such M values cannot work without Alt M1.
Proposal 10: For R2D CP handling, at least Alt M1 CP handling is supported for all M values.
For large M values, Alt M2-1-1 Candidate 3 can be additionally considered.
If Alt M2-1-1 Candidate 3 is additionally specified, it must be required to use it for the applicable M values, so that a single solution exists for all BS which use such an M value.
Proposal 11: One OFDM symbol contains M chips needs to be specified in step 1 in section 4.4 of TR 38.769.
Proposal 12: The exact value of L and exact elements of L samples does not need to be specified in step 2 in section 4.4 of TR 38.769.
Proposal 13: The DFT-points N’ = M * L needs to be specified in step 3 in section 4.4 of TR 38.769.
Proposal 14: The constraint DFT-points N’  X subcarriers of Btx,R2D of the step 3 in section 4.4 of TR 38.769 needs to be specified 
Proposal 15: It is preferable to specify which X elements of N’-points frequency domain signal are mapped to the X subcarriers in step 4 in section 4.4 of TR 38.769. 
Proposal 16: The IDFT-points N of the step 5 in section 4.4 of TR 38.769 needs to be specified, but the number of points N does not.
Proposal 17: The CP insertion (note in section 4.4 of TR 38.769) needs to be specified and it refers to the same operation defined in TS 38.211 section 5.3.1.
Proposal 18: For R2D, the minimum unit of resource allocation, is defined as the chip duration which is equal to one OFDM symbol length without CP divided by M i.e. (1/SCS)/M.
Proposal 19: For R2D padding, the end chip(s) of the padding content shall follow the CP handling method determined in section 2.2.

3GPP TSG RAN WG1 #120bis		R1-2502273
Wuhan, China, April 7th – 11th, 2025

Source:	OPPO
Title:	Discussion on modulation aspects of A-IoT
Agenda item:	9.4.1
Document for:	Discussion and Decision

Conclusion
In this contribution, we discuss the R2D and D2R transmission, including the M values, overlaid OFDM sequence, CP handling, chip rate. Observations and proposals are summarized as following. 

R2D 
Observation 1: More than one transition edges occur in the CP part for M=16,24,32.
In case of M=16, two or three transition edges occur during CP part. When three transition edges occur, the duration between two adjacent transition edge is clearly different.
In case of M=24, two or three transition edges occur during CP part. When three transition edges occur, the duration between two adjacent transition edge is nearly same.
In case of M=32, two or three or four transition edges occur during CP part. 
Proposal 1: The maximum M value is 16.
Observation 2: Overlaid OFDM sequence over OOK chip could help flatten the spectrum, though Ambient IoT device has no capability to detect OFDM sequence, which may or may not have specification impact.
Proposal 2: For the better detection performance of R2D transmission and flatten the spectrum, support overlaid OFDM sequence over OOK-ON chip when OOK-4 waveform generation.
Proposal 3: The method of CP handling discussed in 9.4.1 applies only to R2D data transmission, not to the clock-acquisition part (CAP).
Observation 3: Candidate 1 is essentially same as Method Type 1.
Observation 4: Candidate 2 and 3 is not efficient, only part of OFDM symbol is used to transmit OOK chips.
Observation 5: Candidate 4 and 5 cannot be applied to large M values, e.g. M=16,24,32.
Proposal 4: Support method type1 for CP handling.

D2R
Proposal 5: D2R transmission at least supports the same candidate chip rate/chip length as the R2D transmission.

3GPP TSG RAN WG1 #120bis								R1-2502318
Wuhan, PR China.  7-11 April 2025
Agenda Item 	:	9.4.1
Source 	:	Sony 
Title 	:	Physical channel design aspects for Ambient IoT 
Document for 	:	Discussion 

Conclusion 
This document has considered CP handling for R2D modulation. The following observations and proposals are made:
Observation 1: Subcarrier orthogonality is not strictly required for Release-19 Ambient IoT, but is required for a harmonized air interface that supports device type 2b in a future release.
Proposal 1:  Alt M2-2 CP handling schemes are not pursued in RAN1 since subcarrier orthogonality is required for a harmonized air interface design.
Observation 2: Alt M2-1-1 Candidate 2 for CP handling shows promising BLER performance and robustness to CFO, at least for M=4.
Proposal 2: For CP handling, RAN1 supports Alt M2-1-1 Candidate 2, where a known signal shape in the tail-portion of the R2D signal is copied to the CP.

3GPP TSG RAN WG1 #120bis		R1-2502368
Wuhan, China, April 7th – 11th, 2025 
Agenda item:	9.4.1
Source:	Samsung
Title:	Views on Physical channels design – modulation aspects
Document for:	Discussion and decision

Conclusion
In this contribution, we discussed modulation aspects for physical channel design with the following observations and proposals:
Observation 1: The impact of synchronization inaccuracy should be carefully considered and justified for each alternative of CP handling Method Type 1.
Observation 2: Alt M1-1 is not appropriate for large M values due to sampling offset by not distinguishing LCP and SFO impacts. For instance, with M=32, a typical sampling offset of 25% ~ 90% of code chip length are observed for small payload size of 20 bits, and the sampling offset for large payload size e.g. 96 bits will further increase due to SFO impact, which may severely impacts R2D reception performance.
Observation 3: Alt M1-2 is not appropriate for large M values, e.g. M>8, due to similar chip length and SFO impacts. For instance, with M=12 or M=16, the time length of LCP/NCP and code chip have only 0~2 sampling points difference and become non-distinguishable at receiver side.
Observation 4: The performance of Method type 1 degrades with M value increasing, due to SFO impact and reduced length of code chip. Compared with small value of M=4, the SNR threshold of 10% target BLER with medium value of M=12 increases around 2.5 dB, and with the large value of M=24 and 32 increases around 2.3 and 5 dB, respectively.
Observation 5: The performance of Candidate 3 of Method type 2 is relatively stable under different M values, e.g. ~1.5 dB difference of SNR at target BLER 10% under M={4, 12, 24, 32}, by adding appropriate number of padding chips.
Observation 6: With M value increasing, a non-negligible performance degradation of Method Type 1 is observed, and there exist a performance gap of up to 5 dB between Method Type 1 and Candidate 3 of Method Type 2.
Observation 7: The remaining issues for D2R modulation including 1) a set of supported D2R chip lengths and 2) a signaling for indicating D2R chip length can be discussed in AI 9.4.2 and AI 9.4.4.

Proposal 1: Method type 1 with Alt M1-1, or Alt M1-2, or combination of both alternatives, is not preferred at least for medium and large M values.
Proposal 2: Method Type 2 is preferred at least for medium and large M values. It can be performed by Candidate 3 of Method Type 2 that insert padding chips at both start and end OOK chips or only at end OOK chips.
Proposal 3: CP handling methods is applied depending on M values, as follows:
When M value is equal to or larger than a threshold, Method Type 2 is applied. At least Candidate 3 can be supported to avoid transition edges within CP.
When M value is smaller than a threshold, further study the performance between Method 1 and candidates of Method 2.
FFS value of the threshold, e.g. M=12
Proposal 4: Alt M2-2: Method Type 2 does not retain subcarrier orthogonality, is not supported by Rel-19 A-IoT.
Proposal 5: For the purpose of better solving CP handling issue, the start of R2D transmission should be aligned with NR half-slot boundary to facilitate the determination of underlying OFDM CP length and removal.
Proposal 6: When the generated number of chips for the R2D transmission does not fully occupy the last OFDM symbol, padding is used based on one of the following options:
Option 1) zero-bit padding before CRC attachment
Option 2) zero-chip padding after CRC attachment
Proposal 7: When PRDCH is attached with postamble and padding is performed, the end of postamble is aligned with the next OFDM symbol boundary by padding zero-bits before CRC attachment or zero-chips after CRC attachment. 
Proposal 8: For R2D transmission, the supported M values are limited to {1, [2 and/or 4], 8, 16, 24}.
Proposal 9: For R2D transmission, the M value of a given PRDCH is indicated by preamble associated with the PRDCH.
Proposal 10: M values of PRDCH can be implicitly indicated by OOK chip length of preamble. 
Proposal 11: For PRDCH reception, the M value is identical for both control and data parts.
Proposal 12: For PRDCH chip duration, 
For non-CP part, chip duration is denoted as . 
Within CP part, the R2D OOK chip duration can be larger or less than  after CP insertion by copying code chip(s) in the end of an OFDM symbol. 
Definition of chip duration for R2D preamble is depending on preamble design and discussed separately.



References:
TR 38.769-201, Study on solutions for ambient IoT (Internet of Things)
RP-250896, New Work Item: Solutions for Ambient IoT (Internet of Things) in NR, 3GPP TSG RAN Meeting #107, Incheon, Korea, March 12th-14th, 2025
RAN1 Chair’s Notes, 3GPP TSG RAN WG1 #119, Orlando, US, November 18th – 22nd, 2024.
RAN1 Chair’s Notes, 3GPP TSG RAN WG1 #120, Athens, Greece, February 17th – 21st, 2025.
R1-2502369, Views on Physical channels design – line coding, FEC, CRC, repetition aspects, Samsung

3GPP TSG RAN WG1 #120bis			R1-2502441
Wuhan, China, 07th - 11th April 2025

Agenda Item:	9.4.1
Title:	Discussion on modulation aspects for Ambient IoT
Source:	Xiaomi
Document for:	Decision

Conclusion
In this contribution, we provide the following proposals and observations.
Observation 1: Aligning the R2D beginning with the NR OFDM symbol beginning simplifies the Reader's implementation. However, aligning the R2D end with the NR OFDM symbol end is not energy-efficient for the reader.
Observation 2: From an A-IoT device perspective, aligning the R2D end with the NR OFDM symbol end is not meaningful and may lead to additional listening behavior, which is inefficient.
Observation 3: Method type 1 imposes more stringent requirements on the device, yet it effectively simplifies the complexity on the reader side.
Observation 4: For Alt-M1-1 in R2D CP handling, its performance remains to be further validated due to the challenges in establishing a reliable reference point and potential errors in its underlying assumptions.
Observation 5: For Alt-M1-2 in R2D CP handling, its performance is significantly influenced by the configuration of the M value. To support Alt-M1-2, the device is also required to feature a more precise clock mechanism.
Observation 6: Compared to candidates 1 and 3, candidate 2 for Alt-M2-1-1 exhibits a simpler implementation process and lower standardization complexity.
Observation 7: Candidate 4 and 5 would make device detection more difficult because of distinct M values within an R2D duration.
Observation 8: For Alt-M2-1 in R2D CP handling, various implementations can be adopted, even for the same candidate alternative. Different implementation schemes exert distinct influences on protocols and have varying effects on the implementation in devices and readers.
Observation 9: Alt-M2-2 in R2D CP handling lacks significant advantages and has a considerable impact on other NR transmissions, making it also unsuitable for A-IoT transmission.
Observation 10:When determining the M value, it is crucial to consider the number of information bits to ensure that the device can receive all necessary information within a short time frame, thereby avoiding additional transmission delays.
Observation 11: The evaluation for M values should be grounded in data transmission efficiency and encompass various coverage scenarios. At least three cases must be selected, with resource cost considered as one of the key evaluation criteria.
Observation 12: Case 1-1, case 2-1 and case 3-1 case will not work because its BLER cannot reach 0.1.
Observation 13: In comparison to cases 1-3, 2-2, and 3-2, case 1-2 demonstrates superior overall performance.
Observation 14: The performance of Case 6-2 surpasses that of Case 5-1; although its spectral efficiency is lower, it remains significantly higher than that of Case 4-1.
Observation 15: The bandwidth allocated by the reader for transmitting PRDCH depends on its specific implementation, provided that it meets the minimum bandwidth requirements.
Observation 16: The duration corresponding to the M-value including the CP enhances the performance of R2D by extending the OOK chip length and mitigating ambiguities in OOK chip.
Proposal 1: Support Alt 1 for time domain alignment
The content of padding is up to reader implementation and transparent to A-IoT device.
Proposal 2: Support Option1 for waveform generation
Option 1: Details of DFT-s-OFDM waveform generation is up to reader implementation and is not specified in RAN1.
This does not preclude RAN4 to discuss R2D waveform generation themselves when defining requirement(s), if needed.
Proposal 3: RAN1 at least supports candidate 2 of Alt-M2-1 in R2D CP handling, and the implementation of Alt-M2-1 could be down selected from the following cases.
Case1: Same OOK chip lengths within information part of an OFDM symbol
Case2: Vary OOK chip lengths within information part of an OFDM symbol
Case3: CP as the first OOK chip length within an OFDM symbol
Proposal 4: RAN1 should not support Alt-M2-2 because of adverse effects on co-existence and implementation.
Proposal 5: Support M=16 as the maximum M value for the OOK-4 modulation because of its superior overall performance.
Proposal 6: M values from {1,8,16} should be supported to ensure optimal data transmission efficiency and diverse coverage scenarios.
Proposal 7: Support minimum bandwidth values of 1, 2, 2 PRBs respectively when M=1, 8, 16.
Proposal 8: Support option 1 for R2D chip duration
The chip duration is equal to (1/M) × {OFDM symbol duration without CP}

3GPP TSG RAN WG1 #120bis			R1-2502467
Wuhan, China, 7th – 11th April 2025

Source:	Panasonic
Title: 	A-IoT Physical Channels Design on Modulation Aspects
Agenda Item:		9.4.1
Document for:	Discussion, Decision

Conclusion 
In this contribution, the followings proposals are made: 
Proposal 1: M=24 or higher should be target regardless of CP handling in order to be competitive with UHF RFID.
Proposal 2: R2D is not required to be orthogonal to NR/LTE.
Proposal 3: A unified CP handling should be considered for PRDCH, SIP and CAP.
Proposal 4: The candidate 1 of Method type 2 Alt M2-2 (as per TR 38.769) can be used for PRDCH, SIP and CAP.
Proposal 5: For non-unified CP handling, PRDCH can adopt Method type 1 with removing the position of CP, while SIP and CAP can adopt Method type 2 with low M value.
Proposal 6: A device needs to have pre-knowledge on CW spectrum for energy harvesting. It is desirable that CW spectrum for energy harvesting is same as the one used for PRDCH.

3GPP TSG RAN WG1 #120bis				R1-2502474
Wuhan, China, April 7th – 11th, 2025

Agenda Item:	9.4.1
Source: 	LG Electronics
Title: 	A-IoT PHY layer design – waveform and modulation aspects
Document for:	Discussion and decision

Conclusion
	In this contribution, we shared our views on general aspects of Ambient IoT physical layer design.
Observation 1: Using Method Type 1 Alt 2 alone, Ambient IoT devices can handle the CP for M values up to 8.
Observation 2: Using Method Type 1 Alt 2 combined with Method Type 1 Alt 1, Ambient IoT devices can also handle the CP with larger (>8) M values.
Proposal 1: Support Method Type 1 for CP handling method for Ambient IoT which can support larger (>8) M values with the combination of Alt M1-1 and Alt M1-2, and also provides better resource utilization with less specification impact.
Observation 3: For the candidate schemes for Alt M2-1, resource efficiency needs to be improved.
Observation 4: Using Alt M2-1 combined with Alt M1-1 assuming Manchester coding is used for R2D transmission, Ambient IoT devices can handle CP for M values up to 24.
Observation 5: Even with the combination of Alt M2-1 and Alt M1-1, false rising/falling edge occurs when M=32 because multiple chips are within CP duration.
Proposal 2: Support Option 1 when Method Type 1 is used for CP handling.
Proposal 3: R2D chip duration is indicated in CAP of R-TAS.
Proposal 4: The chip durations for L1 R2D control (if supported) and R2D data are the same.
Proposal 5: The following two alternatives for indicating R2D chip duration in CAP can be considered.
Alt.1) CAP indicating the R2D chip duration directly indicates the R2D chip duration (without partitioning)
Alt.2) CAP indicating the R2D chip duration consists of two parts the combination of which indicates the R2D chip duration
Proposal 6: Support M=32 for the maximum M value.
Proposal 7: Consider defining a basic set of M values (e.g., M = 1, 2, 4, 8) to be supported by all devices in Rel-19. 
FFS defining extended set(s) to be supported based on device capabilities or by devices in later releases.
Proposal 8: For R2D, when the generated number of chips for the R2D transmission does not fully occupy the last OFDM symbol, padding is used.
Specifying the content of padding may be beneficial for AmIoT devices
Proposal 9: Clarification is required on whether the device supports both OOK and BPSK modulation or only one among the available modulation schemes.

3GPP TSG RAN WG1 #120bis	                                                                     R1-2502584
Wuhan, China, April 7th – 11th, 2025

Agenda Item:	9.4.1
Source: 	InterDigital, Inc.
Title:		Discussion on modulation aspects of physical channel design for Ambient IoT
Document for:	Discussion and Decision
1. 

Summary
This contribution has discussed modulation aspects of physical channel design for Ambient IoT. The following are proposed.

Proposal 1: The following steps are specified for R2D waveform generation:
1.   The time domain OOK signal is the M chips of one OFDM symbol.
2.	A chip is represented by L samples where L = N’/M. Chip-1 is represented with a non-zero sequence and chip-0 is represented with a sequenece of zeroes.
3.	N’-point DFT is performed on the samples of one OFDM symbol to obtain the frequency domain signal where N’min = X.
4.	The frequency domain signal obtained by DFT is mapped to the X subcarriers of Btx,R2D. 
5.  IDFT is performed to obtain the time domain signal.

Proposal 2: For R2D, for the OOK-4 modulation for M-chip per OFDM symbol transmission, the maximum M value is 24.

Proposal 3: For R2D, for the OOK-4 modulation for M-chip per OFDM symbol transmission, support at least M = 2, 4, 8, 12, 16, 24.

Proposal 4: The minimum number of RBs associated with each M value is:
Table 3 Minimum number of PRBs


Proposal 5: Support removal of the cyclic prefix at the device without reader side specification.

Conclusion
In this contribution, the following proposals and observations have been made:
Proposal 1: Consider specifying time-domain sequence mapped to ON-chip in Step 2.
Observation 1: Given the same total transmit power, larger values of  outperform smaller  while minimizing both transmission resources and transmission time.
Observation 2: Bandwidths larger than the minimum BW increase performance.
Proposal 2: Allow a maximum  value of 32 to achieve higher data rate than RFID.
Proposal 3: Support a selected number of  values, e.g. 1, 2, 4, 8, 16, 24 and 32.
Proposal 4: The minimum transmission BW  for  should be increased to ensure sufficient performance.
Proposal 5: Define chip duration as  .

3GPP TSG RAN WG1 #120bis		        R1-2502605
Wuhan, China, April 7th – 11th, 2025

Agenda Item:	9.4.1
Source:	Apple
Title:	On modulation aspects for Ambient IoT
Document for:	Discussion/Decision

Conclusion
In this contribution, following observations/proposals have been made related to modulation aspects for ambient IoT:
Observation 1: For future deployment scenarios such as D2T2 and outdoor scenarios, not ensuring subcarrier orthogonality may be problematic with Method Type 2, Alt M2-2, Candidate 1

Observation 2: Method Type 2, Alt M2-1-1, Candidate 1 requires specifying combination of two different solutions depending on value of M and this results in more effort towards an issue for which there are other solutions that may work well for all values of M

Observation 3: Method Type 2, Alt M2-1-1, Candidate 2 has an additional consideration of discussing and defining some know sequence which is added to deal with CP insertion, and this adds unnecessary an additional point of discussion

Observation 4: Among Method Type 2, Alt M2-1-2, Candidate 4 and Method Type 2, Alt M2-1-2, Candidate 5, we fundamentally don’t see any difference and both candidates would result in different M values for certain chips in certain OFDM symbols.

Observation 5: To have comparable date rate as RFID, M = 24 is needed

Observation 6: If M = 32 is considered, there are potential performance issues in accurate determination of chip duration unless excessively long CAP is specified. So overall performance gain for M = 32 may not be significant compared to M = 24

Observation 7: On the minimum value of M, considering M = 1 could require very large overhead/duration for the CAP design, at least a minimum of 3 OFDM symbols, since at least 3 chips are needed in CAP

Observation 8: Chip duration is correlated to CP handling method and for Method Type 1 (including Alt M1-1 and/or Alt M1-2), chip duration is fixed and depends on symbol duration and corresponding value of M



Proposal 1: In RAN1#120bis, at least we should remove candidates that are fundamentally same and/or that require different handling for different M values and/or are not forward compatible. Considering this, at least remove following candidates in RAN1#120bis:
Method Type 2, Alt M2-1-1, Candidate 1
Method Type 2, Alt M2-1-1, Candidate 2
Method Type 2, Alt M2-1-2, Candidate 4
Method Type 2, Alt M2-2, Candidate 1

Proposal 2: As a baseline, Method Type 1 (including Alt M1-1 and/or Alt M1-2) can be considered because of no impact to transmitter side and applicability to all values of M and fixed chip duration
If other candidates impacting transmit sider don’t provide significant gains, then Method Type 1 (including Alt M1-1 and/or Alt M1-2) is adopted.


Proposal 3: Adopt minimum value of M = 2 and maximum value of M = 24 for R2D modulation

Proposal 4: Additional value of M = 6 or 8 and M = 12 or 16 can be considered. Essentially, adopt one of the following candidates set for M values for R2D modulation
Candidate set 1: M = 2, 6, 12 and 24
Candidate set 2: M =2, 8, 16 and 24

Proposal 5: If Method Type 1 (including Alt M1-1 and/or Alt M1-2) is adopted, then chip duration :  

3GPP TSG RAN WG1 #120-bis                               R1-2502671
Wuhan, China, February 7th - 11th, 2025

Source:	Sharp
Title:	Discussion on modulation aspects
Agenda Item:	9.4.1
Document for:	Discussion and Decision

Conclusion
In this contribution, we discuss a few modulation aspects of the A-IoT physical layer channel design, and make the following observations and proposals.
The following should be taken into account for specification of the DFT-s-OFDM waveform for OOK-4 modulation for an R2D transmission:
Synergy with the specification work for the OOK-4 waveform for LP-WUS.
Forward-compatibility with Rel-20 where a Reader can be a 5G UE.
The exact “certain specification of DFT-s-OFDM waveform generation for OOK-4 modulation” is not an urgent issue in the RAN1 Rel-19 A-IoT discussions.
R2D chip length for R2D start-indicator-part (SIP) should correspond to a relatively large value of M, even if the value of M for the corresponding clock-acquisition-part (CAP) and PRDCH is small.
Any part of an R2D transmission (e.g. SIP, CAP, PRDCH, postamble) has a duration in integer number of R2D chips.
Any part of an D2R transmission (e.g. preamble, midamble, PDRCH) has a duration in integer number of D2R chips.
Specification of the DFT-s-OFDM waveform generation for OOK-4 modulation for an R2D transmission can be based on that of the OOK-4 waveform generation in Rel-19 LP-WUS, with a few exceptions including: fixed SCS of 15 kHz, the R2D transmission starting at the start of an (arbitrary) OFDM symbol (up to the Reader), and transmission bandwidth up to the Reader, with the minimum value determined by the value of “M”.
Details can be deferred to specification drafting phase.
For R2D, for the OOK-4 modulation for M-chip per OFDM symbol transmission, the maximum value of M is 16.
For R2D, different chips within one OFDM symbol can correspond to different values of M.
For CP handling which retains subcarrier orthogonality, down-select to Method Type 1.
For R2D, RAN1 to conclude that CP handling Method Type 1 is assumed in the Rel-19 A-IoT WI.
Note 1: use of Alt M1-1 and/or Alt M1-2 is up to device implementation.
Note 2: no specification impact.
R2D chip duration is defined as .
A set of D2R bandwidths (including a number of values as integer multiple of SCS and other values as integer multiple of PRBs) is pre-defined for backscattering transmissions in Rel-19.

3GPP TSG RAN WG1 #120bis		R1-2502690 
Wuhan, China, April 7th – 11th, 2025

Source:	HONOR
Title:	Discussion on Physical channels design for Ambient IoT– modulation aspects
Agenda Item:	9.4.1
Document for:	Discussion and Decision

Conclusions
In this contribution, we provide our views on Physical channels design for Ambient IoT system. The following observations and proposals are given:
Proposal 1: Down select the supported M values, e.g. {2,6,24}.
Proposal 2: For CP handling, support Method Type 1 in which removal of CP at device side.

Observation 1: One CP handling solution means only one method could be selected for both small and large M values.
Observation 2: Alt M2-1 candidate 1 and 2 needs extra processing for devices. 
Observation 3: Alt M2-1 candidate 4 and 5 are not feasible for larger M values, such as M=24. 
Observation 4: Alt M2-1 candidate 3 is feasible for larger M values by inserting more padding chips, but result in degraded spectrum efficiency.
Observation 5: CP handling of Alt M2-2 does not retain subcarrier orthogonality which would cause interference to NR system, and increase the complexity of the transmitter. 
Observation 6: A-IoT devices can support Method Type 1 by implementing Alt M1-1 or Alt M1-2 depending on M value indicated by preamble without standard effort.

3GPP TSG RAN WG1 #120bis		R1-2502704
Wuhan, China, April 7th – 11th, 2025
Agenda item:	9.4.1
Source: 	MediaTek Inc.
Title: 	Discussion on A-IoT physical modulation aspects
Document for:	Discussion/Decision

Conclusion
 Observation 1: Regarding the maximum M value of OOK-4 modulation, either 24 or 32 can achieve a comparable or even higher data rate than RFID.
Observation 2: Regarding the maximum M value of OOK-4 modulation as 24 or 32, the corresponding performance and spectrum efficiency need to be considered.
Observation 3: For the M value detection, assuming 10% residual SFO at the device, any two adjacent M values within {2, 4, 6, 8, 12, 16, 24, 32} suffer the possibility of false detection.
Proposal 1: As per TR 38.769, for the OFDM-based OOK-4 waveform, from the reader perspective, when the generated number of chips for the R2D transmission does not fully occupy the last OFDM symbol, padding is used to align with the boundary of the last NR OFDM symbol for in-band operation. The content of padding is up to reader implementation and transparent to A-IoT device, i.e., Alt 1.
Proposal 2: Support 24 as the maximum M value of OOK-4 modulation. 
Proposal 3: Support 1 as the minimum M value of OOK-4 modulation.
Proposal 4: In addition to the maximum and minimum M values, the selection of other M values from {2, 4, 6, 8, 12, 16, 24} should be non-adjacent.
Proposal 5: Regarding CP handling, Method type 1 is supported with:
Alt M1-2 for OOK-4  
Alt M1-1 for OOK-4 

3GPP TSG RAN WG1 #120bis							         R1-2502766
Wuhan, China, April 7th – 11th, 2025

Source:	NTT DOCOMO, INC.
Title:	Discussion on modulation aspects of physical channel design for Ambient IoT
Agenda Item:	9.4.1
Document for: 	Discussion and Decision

Conclusion
In this contribution, we discussed modulation aspects on physical channel design for A-IoT device. Based on the discussion, we made following observations and proposals.
R2D
Observation 1: For PRDCH, to achieve more than one hundred kbps data rate, at least M=16 should be supported considering that Manchester coding is applied.
Observation 2: For PRDCH, for CP handling Method 1, at least Alt.1 can work for any M value assuming that device can identify the CP location.
Observation 3: For PRDCH, for CP handling Method 2 Alt.1, the characteristic of Manchester coding such as transition edge at the middle of codeword is not preserved for last chips when M=16 or larger.
Observation 4: For PRDCH, CP handling Method 2 Alt.1-1 may not work for M=2 assuming the codeword of Manchester coding as; bit 0→chips{10}, bit 1→chips{01}.
Observation 5: For PRDCH, CP handling Method 2 Alt.2 can work for any M value.

Proposal 1: For CP handling for PRDCH, one of CP handling Method type 1, Method type 2 Alt.1 or Method type Alt.2 should be down-selected regardless of M value for OOK-4.
Proposal 2: For PRDCH, support CP handling Method Type 1.
Note: CP handling for R-TAS should be discussed separately from that for PRDCH.
Proposal 3: Discuss how a device determines the OFDM symbol boundary for CP handling of PRDCH reception.
E.g., Counting of samples from start/end of R-SIP/CAP of preamble, i.e., start/end of R-SIP/CAP is aligned with OFDM symbol boundary.
Proposal 4: For PRDCH, discuss whether the impact on the difference of short CP symbol and long CP symbol, i.e., 16κ per 0.5ms, is negligible or not.
Proposal 5: For PRDCH, the maximum value on M value of OOK-4 should be 16 or 24.
Proposal 6: For PRDCH, at least 1, 2, 4, 8 and 16 should be supported for M value of OOK-4
Note: M values for R2D CAP should be discussed separately from that for PRDCH
Proposal 7: For R2D, support minimum transmission bandwidth for each M value of OOK-4 which is captured in table 6.1.1.4-1 in TR 38.769 as baseline.
Note: Whether larger value should be specified for any M value can be discussed in RAN4.
Proposal 8: From RAN1 perspective, it is not necessary to specify the detailed DFT-s-OFDM waveform generation procedure for OOK-4 modulation.
Note: Any requirement on R2D waveform generation can be specified in RAN4.
Proposal 9: For PRDCH, one chip length corresponds to (one OFDM symbol excluding CP)/M if CP handling Method 1 or Method 2 combined with Method 1 is supported, otherwise, one chip length corresponds to (one OFDM symbol including CP)/M.
Proposal 10: For PRDCH resource allocation unit in time, discuss the following alternatives if indication of time domain resource for R2D is supported.
Alt.1: Manchester code codeword, e.g., two chips.
Alt.2: OFDM symbol, i.e., M chips.

D2R
Proposal 11: The set of D2R information bit length Tb is defined and the candidate values should consider the following aspects;
Minimum Tb length should consider achievable maximum baseband data rate for D2R.
Note: maximum Tb length can be determined based on the supported D2R chip duration and R.
Proposal 12: For the set of predefined D2R chip durations, the candidate values should consider the following aspects;
Transmission bandwidth should be integer multiple of 15 kHz.
Minimum chip length should consider the device sampling clock frequency.
Proposal 13: For the predefined set of D2R chip duration values, the maximum chip length can be 133.33μs.

3GPP TSG RAN WG1 Meeting #120-bis    	            R1-2502839
Wuhan, China, Apr 7 – 11, 2025

Source: Qualcomm Incorporated
Title: Physical channel design - modulation aspects
Agenda Item:	9.4.1
Document for: 	Discussion and Decision

Conclusion
For CP handling Method M1-1, poor clock accuracy affects the identification of CP location and removal of CP and degrade performance.
For CP handling Method M1-1, during CAP detection, device may need to adjust its internal clock or adjust internal clock tick threshold for the detected chip duration (i.e., M value) to make accurate removal of CP duration.
CP handling Method M1-1 can support up to M=24 in TDL-A fading channel.
CP handling Method M1-1 cannot support M=32 in TDL-A fading channel.
For CP handling method M1-2, device cannot differentiate chips in CP and chips in actual OFDM symbols, which makes M1-2 fail in large M>=12.
For CP handling scheme M2-1-2, the duration of chip before/after CP is dependent on the last information bit of the OFDM symbol.
For CP handling scheme M2-1-2, chip duration is not uniform.
M2-1-2 Candidate 4 require device to detect two M values and measurements.
For CP handling scheme M2-1-2, non-uniform chip duration or edge-to-edge duration makes it almost impossible to use embedded clocking information.
For CP handling scheme M2-1-2, varying center point of OOK bit makes it difficult to apply other detection schemes such as energy detection.
CP handling method M2-1-2 is not applicable to the case that CP duration is longer than chip duration, e.g., M>12.

CP handling method M2-2 can support M=24 in TDL-A fading channel.
CP handling method M2-2 cannot support M=32 in TDL-A fading channel.
The benefit of orthogonal CP handling scheme is valid only when outdoor NR gNB is synchronized with indoor A-IoT gNB.
In FDD spectrum, gNB time synchronization is not strongly required.
If outdoor gNB and indoor gNB are not synchronized, then, even orthogonal CP handling scheme introduces ICI. Thus, there is no strong benefit of using orthogonal CP handling scheme in FDD spectrum.
CP handling Method Type 1 is implementation-based solutions.
CP handling method type 1 have burden mostly at device side. CP handling Method Type 2 has burden mostly at reader side.
Choice of CP handling scheme decides OOK chip duration.
For CP handling scheme, RAN1 to prioritize CP handling methods of Alt M2-2 and Alt M1-1.
RFID reader can use any information data rate between 26.6kbps and 128kbps.
In R2D with Manchester coding, M=4 (28kbps) to M=16(112bkps) provides similar range as that supported by RFID.
Device can use identified M value in detecting OFDM symbol boundary, CP handling (if any), and device clock calibration.
A-IoT device should be able to identify indicated M value based on CAP.
M = 2 or M=4 can be considered as minimum M value for PRDCH data.
In CAP M detection, Tx M=6 could be detected as M=8 with high false detection probability of 0.106.
In CAP M detection, Tx M=12 could be detected as M=16 with high false detection probability of 0.09.
Support only one of M=6 and 8.
Support only one of M=12 and 16
CP handling Method M1-1 can support M=24 in TDL-A fading channel.
CP handling Method M1-1 cannot support M=32 in TDL-A fading channel.
CP handling method M2-2 can support M=24 in TDL-A fading channel.
CP handling method M2-2 cannot support M=32 in TDL-A fading channel.
CP handling method M2-2 performs better than M1-1.
For PRDCH, support a set of M values = {2, 4, 8, 12, 24}.
For CAP, support a set of M values = {1, 2, 4, 6, 12}.
Support following mapping between CAP M value and M value of PRDCH part.

Limiting the set of M values can improve M value detection performance.

In terms of RFID FL data rate, RFID BL data rate range is given as .
The wide range of BL data rate implies that FL and BL data rate can be almost independently chosen by reader.

3GPP TSG RAN WG1 Meeting #120-bis	                                                           R1-2502898
Wuhan, China, April 7th – 11th, 2025

Agenda Item:  	9.4.1
Source:             	Fraunhofer IIS, Fraunhofer HHI
Title:                  	Discussion on modulation aspects for Ambient-IOT
Document for: 		Decision

Conclusion
The following proposals are made from the above discussion:
Proposal 1: For the specification for OOK-4 waveform generation based on DFT-s-OFDM, the following need not be specified:
The bit-spreading sequence(s) that can be used, 
Possible waveform shaping techniques before and after DFT-spreading. 
Proposal 2: For the specification for OOK-4 waveform generation based on DFT-s-OFDM, the following shall be applicable: 
N’ shall be the same as the size of the vector that is applied with the DFT. 
The value of N from legacy OFDM generation may be retained. 
Proposal 3: For the specification for OOK-4 waveform generation based on DFT-s-OFDM, the following details shall be included: 
Relationship between the parameters M, L, N’ and X
Proposal 4: RAN1 shall study the use and applicability of specifying parameters regarding OOK pulse shape and form to be obtained from OOK-4 DFT-s-OFDM waveform generation. 

3GPP TSG RAN WG1 #120-bis		R1-2502905
 Wuhan, China, April 7th – 11th, 2025  

Agenda Item:	9.4.1
Source:	Lenovo
Title:	Discussion on modulation aspects for Ambient IoT physical channels 
Document for:	Discussion and Decision 

Conclusion
In this contribution, we focus on modulation aspects of Ambient IoT physical channels with following proposals:
Proposal 1: For CP handling, support removing CP at device without impact on transmitter. 
      FFS: Down scope to AltM1-1 or AltM1-2 for determining the CP length.
Proposal 2: For R2D transmission, limit the maximum M value to (M < 16)
Proposal 3: For R2D transmission consider L values based on M and number of PRBs, L= 12*PRBs /M.

3GPP TSG RAN WG1 Meeting#120bis		R1-2502927
Wuhan, China, April 7th – 11th, 2025

Source:	WILUS Inc.
Title:	Discussion on modulation aspects of physical channels design for Ambient IoT
Agenda item:	9.4.1
Document for:	Discussion/Decision

Conclusion
In this contribution, we discussed CP handling, padding for Ambient IoT and summarize our views as the following:
Proposal 1: Support CP Handling Alt.M1-1 with a combination of Alt. M2-1-1 inserting padding chips only at end OOK chips depending on M values.
Alt.M1-1 if M =<12
Alt.M2-1-1 inserting padding chips only at end OOK chips if M > 12

Proposal 2: In R2D transmission, if padding is required at the last OFDM symbol, two alternative handling methods are proposed:
If the PRDCH is immediately followed by a postamble:
Fill padding chips with all ‘0’s or ‘1’s.
Chip level should match either the first chip of the current OFDM symbol or the last chip of the previous symbol.
If the PRDCH is not immediately followed by a postamble:
Alt 1: Fill padding chips with all ‘0’s or ‘1s, matching the first chip of the current OFDM symbol.
Alt 2: If there are more than four paddings chips to be filled, the first two padding chips match the last data chip. Remaining chips use ‘01’/‘10’ pattern, with the last chip matches the first chip of the OFDM symbol. 

TDoc file unavailable

3GPP TSG RAN-WG1 Meeting #120bis			R1-2502992
Wuhan, China, 07-11 April, 2025

Source: 	Moderator (Huawei)
Title:	FL summary #1 for Ambient IoT: “9.4.1 Physical channels design – modulation aspects”
Document for:	Discussion and decision
Agenda item:	9.4.1

Conclusion
For CW waveform, A-IoT WID RP-243326 and TR 38.769 already give us:
It is left up to specification drafting phase how to capture the following.


Agreement
For R2D, for the OOK-4 modulation for M-chip per OFDM symbol transmission:
The maximum M value is no less than 16, and to be down-selected from 32, 24 or 16 at RAN1#120bis.

Agreement
For CP handling which retains subcarrier orthogonality, the candidate methods are clarified as follows:
For Method Type 1 (as per TR 38.769)
For purposes of evaluating Method Type 1, it is assumed the device is aware of or determines the boundary of an OFDM symbol using the R2D timing acquisition signal (R-TAS), at least for Alt M1-1
Detail is up to R2D preamble design
Using Alt M1-1 and/or Alt M1-2 is up to device implementation as per TR 38.769
FFS: applicable M values
For Method Type 2 Alt M2-1 (as per TR 38.769)
Detail design of Alt M2-1-1 or Alt M2-1-2 from the followings:
Alt M2-1-1
Candidate 1: 
For M less than 6: Use Alt M1-1
For M >=6: circular shift of the M chips’ samples before CP insertion at the reader side in addition to bit/chip skipping/dropping at the device side after bit decoding
Candidate 2:
Where a known signal shape in the tail-portion of the R2D signal is copied to the CP.
Candidate 3:
Insert padding chips at both start and end OOK chips or only at end OOK chips
Alt M2-1-2
Candidate 4:
Change  case to  case by changing OOK-4 M value to adjacent M +/- 1 for OFDM symbol n.
where a, b and c is the last chip of symbol n, first chip of symbol n+1 and last chip of symbol n+1.
Candidate 5:
Odd number of chips of PRDCH is transmitted in the first OFDM symbols, while M value is kept unchanged during the remaining OFDM symbols.
FFS: applicable M values with or without combination of Method Type 1

For CP handling which does not retain subcarrier orthogonality, the candidate methods are clarified as follows:
For Method Type 2 Alt M2-2 (as per TR 38.769)
FFS detail design of Alt M2-2 from the following:
Candidate 1:
A duration of OFDM symbol with CP has integer number of OOK symbols. CP is not copied from the end of OFDM symbol. Total 14 OFDM symbols fit in a slot.

For further down-selection among candidate methods which retains subcarrier orthogonality, companies are encouraged to provide evaluation (including overhead) among the followings to RAN1#120bis:
Method Type 1
Method Type 2 Alt M2-1-1 with or without combination of Method Type 1
Method Type 2 Alt M2-1-2 with or without combination of Method Type 1
Note: The evaluation for CP handling at least includes a M value from {2,4,6,8}, a M value from {12, 16}, and a M value from {24, 32}
Note: In this evaluation, the transmit power of the ON chips remains constant

Agreement
For D2R BPSK modulation
After line coding or square wave, chip “1” is mapped to the real-valued modulation symbol {1} and chip “0” is mapped to the real-valued modulation symbol {-1} in the baseband.

Agreement
For D2R OOK modulation
After line coding or square wave, chip “1” is mapped to the real-valued modulation symbol {1} and chip “0” is mapped to the real-valued modulation symbol {0} in baseband.

Agreement
For R2D transmission, from reader perspective, DFT-s-OFDM waveform is supported for OOK-4 modulation. For the details of DFT-s-OFDM waveform generation of OOK-4 modulation:
Certain specification of DFT-s-OFDM waveform generation for OOK-4 modulation is needed in RAN1.
FFS: identify what potential details of 5 steps in section 4.4 of TR 38.769 needs to be specified.
This does not preclude RAN4 to discuss R2D waveform generation themselves when defining requirement(s), if needed.

RAN1#120bis (Wuhan)

Feature lead guidance for RAN1#121 Malta

3GPP TSG RAN-WG1 Meeting #120bis			R1-2502993
Wuhan, China, 07-11 April, 2025

Source: 	Moderator (Huawei)
Title:	FL summary #2 for Ambient IoT: “9.4.1 Physical channels design – modulation aspects”
Document for:	Discussion and decision
Agenda item:	9.4.1

Conclusion
For CW waveform, A-IoT WID RP-243326 and TR 38.769 already give us:
It is left up to specification drafting phase how to capture the following.


Agreement
For R2D, for the OOK-4 modulation for M-chip per OFDM symbol transmission:
The maximum M value is no less than 16, and to be down-selected from 32, 24 or 16 at RAN1#120bis.

Agreement
For CP handling which retains subcarrier orthogonality, the candidate methods are clarified as follows:
For Method Type 1 (as per TR 38.769)
For purposes of evaluating Method Type 1, it is assumed the device is aware of or determines the boundary of an OFDM symbol using the R2D timing acquisition signal (R-TAS), at least for Alt M1-1
Detail is up to R2D preamble design
Using Alt M1-1 and/or Alt M1-2 is up to device implementation as per TR 38.769
FFS: applicable M values
For Method Type 2 Alt M2-1 (as per TR 38.769)
Detail design of Alt M2-1-1 or Alt M2-1-2 from the followings:
Alt M2-1-1
Candidate 1: 
For M less than 6: Use Alt M1-1
For M >=6: circular shift of the M chips’ samples before CP insertion at the reader side in addition to bit/chip skipping/dropping at the device side after bit decoding
Candidate 2:
Where a known signal shape in the tail-portion of the R2D signal is copied to the CP.
Candidate 3:
Insert padding chips at both start and end OOK chips or only at end OOK chips
Alt M2-1-2
Candidate 4:
Change  case to  case by changing OOK-4 M value to adjacent M +/- 1 for OFDM symbol n.
where a, b and c is the last chip of symbol n, first chip of symbol n+1 and last chip of symbol n+1.
Candidate 5:
Odd number of chips of PRDCH is transmitted in the first OFDM symbols, while M value is kept unchanged during the remaining OFDM symbols.
FFS: applicable M values with or without combination of Method Type 1

For CP handling which does not retain subcarrier orthogonality, the candidate methods are clarified as follows:
For Method Type 2 Alt M2-2 (as per TR 38.769)
FFS detail design of Alt M2-2 from the following:
Candidate 1:
A duration of OFDM symbol with CP has integer number of OOK symbols. CP is not copied from the end of OFDM symbol. Total 14 OFDM symbols fit in a slot.

For further down-selection among candidate methods which retains subcarrier orthogonality, companies are encouraged to provide evaluation (including overhead) among the followings to RAN1#120bis:
Method Type 1
Method Type 2 Alt M2-1-1 with or without combination of Method Type 1
Method Type 2 Alt M2-1-2 with or without combination of Method Type 1
Note: The evaluation for CP handling at least includes a M value from {2,4,6,8}, a M value from {12, 16}, and a M value from {24, 32}
Note: In this evaluation, the transmit power of the ON chips remains constant

Agreement
For D2R BPSK modulation
After line coding or square wave, chip “1” is mapped to the real-valued modulation symbol {1} and chip “0” is mapped to the real-valued modulation symbol {-1} in the baseband.

Agreement
For D2R OOK modulation
After line coding or square wave, chip “1” is mapped to the real-valued modulation symbol {1} and chip “0” is mapped to the real-valued modulation symbol {0} in baseband.

Agreement
For R2D transmission, from reader perspective, DFT-s-OFDM waveform is supported for OOK-4 modulation. For the details of DFT-s-OFDM waveform generation of OOK-4 modulation:
Certain specification of DFT-s-OFDM waveform generation for OOK-4 modulation is needed in RAN1.
FFS: identify what potential details of 5 steps in section 4.4 of TR 38.769 needs to be specified.
This does not preclude RAN4 to discuss R2D waveform generation themselves when defining requirement(s), if needed.

RAN1#120bis (Wuhan)

Agreement
For R2D, for the OOK-4 modulation for M-chip per OFDM symbol transmission, the maximum M value is 24.
RAN1 will further determine the set of M values up to the maximum M value.
The maximum M value applicable to the PRDCH is 24
The maximum M value applicable to the R-TAS is not larger than 24


Agreement
For R2D, at least for PRDCH, the set of M values is {2, 6, 12, 24}
FFS: whether/how CAP use same M values set as PRDCH

Feature lead guidance for RAN1#121 Malta

3GPP TSG RAN-WG1 Meeting #120bis			R1-2502994
Wuhan, China, 07-11 April, 2025

Source: 	Moderator (Huawei)
Title:	FL summary #3 for Ambient IoT: “9.4.1 Physical channels design – modulation aspects”
Document for:	Discussion and decision
Agenda item:	9.4.1

Conclusion
For CW waveform, A-IoT WID RP-243326 and TR 38.769 already give us:
It is left up to specification drafting phase how to capture the following.


Agreement
For R2D, for the OOK-4 modulation for M-chip per OFDM symbol transmission:
The maximum M value is no less than 16, and to be down-selected from 32, 24 or 16 at RAN1#120bis.

Agreement
For CP handling which retains subcarrier orthogonality, the candidate methods are clarified as follows:
For Method Type 1 (as per TR 38.769)
For purposes of evaluating Method Type 1, it is assumed the device is aware of or determines the boundary of an OFDM symbol using the R2D timing acquisition signal (R-TAS), at least for Alt M1-1
Detail is up to R2D preamble design
Using Alt M1-1 and/or Alt M1-2 is up to device implementation as per TR 38.769
FFS: applicable M values
For Method Type 2 Alt M2-1 (as per TR 38.769)
Detail design of Alt M2-1-1 or Alt M2-1-2 from the followings:
Alt M2-1-1
Candidate 1: 
For M less than 6: Use Alt M1-1
For M >=6: circular shift of the M chips’ samples before CP insertion at the reader side in addition to bit/chip skipping/dropping at the device side after bit decoding
Candidate 2:
Where a known signal shape in the tail-portion of the R2D signal is copied to the CP.
Candidate 3:
Insert padding chips at both start and end OOK chips or only at end OOK chips
Alt M2-1-2
Candidate 4:
Change  case to  case by changing OOK-4 M value to adjacent M +/- 1 for OFDM symbol n.
where a, b and c is the last chip of symbol n, first chip of symbol n+1 and last chip of symbol n+1.
Candidate 5:
Odd number of chips of PRDCH is transmitted in the first OFDM symbols, while M value is kept unchanged during the remaining OFDM symbols.
FFS: applicable M values with or without combination of Method Type 1

For CP handling which does not retain subcarrier orthogonality, the candidate methods are clarified as follows:
For Method Type 2 Alt M2-2 (as per TR 38.769)
FFS detail design of Alt M2-2 from the following:
Candidate 1:
A duration of OFDM symbol with CP has integer number of OOK symbols. CP is not copied from the end of OFDM symbol. Total 14 OFDM symbols fit in a slot.

For further down-selection among candidate methods which retains subcarrier orthogonality, companies are encouraged to provide evaluation (including overhead) among the followings to RAN1#120bis:
Method Type 1
Method Type 2 Alt M2-1-1 with or without combination of Method Type 1
Method Type 2 Alt M2-1-2 with or without combination of Method Type 1
Note: The evaluation for CP handling at least includes a M value from {2,4,6,8}, a M value from {12, 16}, and a M value from {24, 32}
Note: In this evaluation, the transmit power of the ON chips remains constant

Agreement
For D2R BPSK modulation
After line coding or square wave, chip “1” is mapped to the real-valued modulation symbol {1} and chip “0” is mapped to the real-valued modulation symbol {-1} in the baseband.

Agreement
For D2R OOK modulation
After line coding or square wave, chip “1” is mapped to the real-valued modulation symbol {1} and chip “0” is mapped to the real-valued modulation symbol {0} in baseband.

Agreement
For R2D transmission, from reader perspective, DFT-s-OFDM waveform is supported for OOK-4 modulation. For the details of DFT-s-OFDM waveform generation of OOK-4 modulation:
Certain specification of DFT-s-OFDM waveform generation for OOK-4 modulation is needed in RAN1.
FFS: identify what potential details of 5 steps in section 4.4 of TR 38.769 needs to be specified.
This does not preclude RAN4 to discuss R2D waveform generation themselves when defining requirement(s), if needed.

RAN1#120bis (Wuhan)

Agreement
For R2D, for the OOK-4 modulation for M-chip per OFDM symbol transmission, the maximum M value is 24.
RAN1 will further determine the set of M values up to the maximum M value.
The maximum M value applicable to the PRDCH is 24
The maximum M value applicable to the R-TAS is not larger than 24


Agreement
For R2D, at least for PRDCH, the set of M values is {2, 6, 12, 24}
FFS: whether/how CAP use same M values set as PRDCH

Agreement
For further down-selection among CP handing which retains subcarrier orthogonality, at least for PRDCH, at least Method Type 1 is supported
For supported M values <= 12
RAN1 will not further pursue additional CP handling design
For supported M values > 12
RAN1 will further down-select one from the followings
Option 1: Candidate 3 of M2-1-1 (as per agreements from RAN1#120)
Insert padding chips only at the end OOK chips of OFDM symbol
Option 3: RAN1 will not further pursue additional CP handling design

Agreement
Proposal: For Ambient IoT, RAN1 clarify that the definition of PRB is same as NR in TS 38.211.

Feature lead guidance for RAN1#121 Malta

3GPP TSG RAN-WG1 Meeting #120bis			R1-2503109
Wuhan, China, 07-11 April, 2025

Source: 	Moderator (Huawei)
Title:	FL summary #4 for Ambient IoT: “9.4.1 Physical channels design – modulation aspects”
Document for:	Discussion and decision
Agenda item:	9.4.1

Conclusion
For CW waveform, A-IoT WID RP-243326 and TR 38.769 already give us:
It is left up to specification drafting phase how to capture the following.


Agreement
For R2D, for the OOK-4 modulation for M-chip per OFDM symbol transmission:
The maximum M value is no less than 16, and to be down-selected from 32, 24 or 16 at RAN1#120bis.

Agreement
For CP handling which retains subcarrier orthogonality, the candidate methods are clarified as follows:
For Method Type 1 (as per TR 38.769)
For purposes of evaluating Method Type 1, it is assumed the device is aware of or determines the boundary of an OFDM symbol using the R2D timing acquisition signal (R-TAS), at least for Alt M1-1
Detail is up to R2D preamble design
Using Alt M1-1 and/or Alt M1-2 is up to device implementation as per TR 38.769
FFS: applicable M values
For Method Type 2 Alt M2-1 (as per TR 38.769)
Detail design of Alt M2-1-1 or Alt M2-1-2 from the followings:
Alt M2-1-1
Candidate 1: 
For M less than 6: Use Alt M1-1
For M >=6: circular shift of the M chips’ samples before CP insertion at the reader side in addition to bit/chip skipping/dropping at the device side after bit decoding
Candidate 2:
Where a known signal shape in the tail-portion of the R2D signal is copied to the CP.
Candidate 3:
Insert padding chips at both start and end OOK chips or only at end OOK chips
Alt M2-1-2
Candidate 4:
Change  case to  case by changing OOK-4 M value to adjacent M +/- 1 for OFDM symbol n.
where a, b and c is the last chip of symbol n, first chip of symbol n+1 and last chip of symbol n+1.
Candidate 5:
Odd number of chips of PRDCH is transmitted in the first OFDM symbols, while M value is kept unchanged during the remaining OFDM symbols.
FFS: applicable M values with or without combination of Method Type 1

For CP handling which does not retain subcarrier orthogonality, the candidate methods are clarified as follows:
For Method Type 2 Alt M2-2 (as per TR 38.769)
FFS detail design of Alt M2-2 from the following:
Candidate 1:
A duration of OFDM symbol with CP has integer number of OOK symbols. CP is not copied from the end of OFDM symbol. Total 14 OFDM symbols fit in a slot.

For further down-selection among candidate methods which retains subcarrier orthogonality, companies are encouraged to provide evaluation (including overhead) among the followings to RAN1#120bis:
Method Type 1
Method Type 2 Alt M2-1-1 with or without combination of Method Type 1
Method Type 2 Alt M2-1-2 with or without combination of Method Type 1
Note: The evaluation for CP handling at least includes a M value from {2,4,6,8}, a M value from {12, 16}, and a M value from {24, 32}
Note: In this evaluation, the transmit power of the ON chips remains constant

Agreement
For D2R BPSK modulation
After line coding or square wave, chip “1” is mapped to the real-valued modulation symbol {1} and chip “0” is mapped to the real-valued modulation symbol {-1} in the baseband.

Agreement
For D2R OOK modulation
After line coding or square wave, chip “1” is mapped to the real-valued modulation symbol {1} and chip “0” is mapped to the real-valued modulation symbol {0} in baseband.

Agreement
For R2D transmission, from reader perspective, DFT-s-OFDM waveform is supported for OOK-4 modulation. For the details of DFT-s-OFDM waveform generation of OOK-4 modulation:
Certain specification of DFT-s-OFDM waveform generation for OOK-4 modulation is needed in RAN1.
FFS: identify what potential details of 5 steps in section 4.4 of TR 38.769 needs to be specified.
This does not preclude RAN4 to discuss R2D waveform generation themselves when defining requirement(s), if needed.

RAN1#120bis (Wuhan)

Agreement
For R2D, for the OOK-4 modulation for M-chip per OFDM symbol transmission, the maximum M value is 24.
RAN1 will further determine the set of M values up to the maximum M value.
The maximum M value applicable to the PRDCH is 24
The maximum M value applicable to the R-TAS is not larger than 24


Agreement
For R2D, at least for PRDCH, the set of M values is {2, 6, 12, 24}
FFS: whether/how CAP use same M values set as PRDCH

Agreement
For further down-selection among CP handing which retains subcarrier orthogonality, at least for PRDCH, at least Method Type 1 is supported
For supported M values <= 12
RAN1 will not further pursue additional CP handling design
For supported M values > 12
RAN1 will further down-select one from the followings
Option 1: Candidate 3 of M2-1-1 (as per agreements from RAN1#120)
Insert padding chips only at the end OOK chips of OFDM symbol
Option 3: RAN1 will not further pursue additional CP handling design

Agreement
Proposal: For Ambient IoT, RAN1 clarify that the definition of PRB is same as NR in TS 38.211.

Agreement
For the below agreement, further update on the followings
Agreement
For further down-selection among CP handing which retains subcarrier orthogonality, at least for PRDCH, at least Method Type 1 is supported
For supported M values <= 12
RAN1 will not further pursue additional CP handling design
For supported M values > 12
RAN1 will further down-select one from the followings
Option 1: Candidate 3 of M2-1-1 (as per agreements from RAN1#120)
Insert padding chips only at the end OOK chips of OFDM symbol
The last 2 out of M OOK chips at the end of an OFDM symbol are always ‘ON’
Option 3: RAN1 will not further pursue additional CP handling design

Agreement
When the generated number of chips for the R2D transmission does not fully occupy the last OFDM symbol, padding is used.
Padding is to be down-selected among the following alternatives:
Alt 1a: The content of padding is up to reader implementation and transparent to device.
Note: the timeline determination of any timing relationship refers to the end of padding.
Note: it implies the device should be aware of the duration of padding or the last OFDM symbol boundary by implementation. FFS how accurately the device can be aware.
Alt 1b: The content of padding is up to reader implementation and transparent to device.
Note: the timeline determination of any timing relationship refers to the start of the padding.
Alt 2: Specify the detailed content of padding
Note: the timeline determination of any timing relationship refers to the end of padding.
Note: the end chip(s) of the padding content shall follow the CP handling solution determined in RAN1, and may be affected by other agreements.

Agreement
From reader perspective, for the needed certain specification of DFT-s-OFDM waveform generation for OOK-4 modulation:
An example is provided below, which does not presume any specific reader implementation:
Step 1: The time domain OOK signal is the M chips of one OFDM symbol
The specification only needs to reflect that one OFDM symbol contains M chips
Step 2: A chip is represented (e.g. upsampled) by L samples
The specification only needs to reflect that one chip contains L samples as the input to N’-points DFT
Step 3: An N’-points DFT is performed on the samples of one OFDM symbol to obtain the frequency domain signal.
The specification only needs to reflect that there is an N’-points DFT operation, where N’ = M*L
Step 4: Map the frequency domain signal obtained by N’-points DFT to the X subcarriers of Btx,R2D
The specification only needs to reflect that N’ >= X, where X is corresponding to the Btx,R2D
Step 5: An N-points IDFT is performed to obtain the time domain signal.
The specification only needs to reflect that there is an N-points IDFT operation
Note: other examples were provided in contributions to RAN1#120bis, e.g. in annex 2 of R1-2502160
From the example above, some normative specification related to at least step 1 and step 5 are needed.
Note: RAN1 to consider whether an information annex could describe other steps
Note: some normative RAN1 specification text about waveform is assumed to be needed for RAN4 requirements definition
Note: the specification also needs to reflect the timing of the CP insertion operation.

Feature lead guidance for RAN1#121 Malta
Based on the discussion and agreements achieved in RAN1#120bis, from FL’s perspective, companies are encouraged to contribute at least on the followings for next meeting RAN1#121:
R2D CP handling
Down-select CP handling at least for PRDCH.
Determine whether/how CP handling for R-TAS, if any.
R2D chip duration
Define the time duration of R2D chip.
R2D numerology
Down-select the content of padding with taking into consideration of the timing reference point (the start or the end of the padding) for timeline determination.
R2D waveform
Continue to discuss the needed certain specification of DFT-s-OFDM waveform generation for OOK-4 modulation, according to agreements achieved in previous meetings.

Others can also be discussed if companies find anything missing for specification of modulation aspects.