Conclusion
In this section, our proposals on UE features for LTE based 5G broadcast operation in Rel-19 are summarized.
Proposal 1: Time-interleaving and frequency-interleaving should be defined as two independent UE FGs.
Observation 1: For FG about time-interleaving transmission of LTE based 5G broadcast, at least the following aspects should be included:
Reception of one TB which is mapped to N non-consecutive subframes;
Two transmissions of the same TB are separated by (M-1) subframes;
TBS determination for the scaled TB;
Determination of the starting point for reading from the circular buffer (k0) for each subframe.
Observation 2: For FG about frequency-interleaving transmission of LTE based 5G broadcast, at least the following aspects should be included:
Data symbols in an OFDM symbol are written column-wise;
Data symbols in an OFDM symbol are read row-wise;
The granularity of frequency interleaving is one RE.
Proposal 2: Define the following UE FGs for Rel-19 LTE based 5G broadcast.
|
3GPP TSG-RAN WG1 Meeting #120bis R1-2502252
Wuhan, China, April 7th – 11th, 2025
Agenda Item: 9.15.13
Source: Huawei, HiSilicon
Title: UE features for LTE based 5G broadcast Phase 2
Document for: Discussion and Decision
|
3GPP TSG RAN WG1 #120-bis R1-2502874
Wuhan, China, April 7th – 11th, 2025
Agenda item: 9.15.13
Source: Qualcomm Incorporated
Title: UE features for LTE-based 5G Broadcast Phase 2
Document for: Discussion and Decision
Features for time and frequency interleaving
In this contribution, we describe the UE features we need for time and frequency interleaving in 5G Broadcast, along with the necessary details that are relevant for these features.
First off, we note that, frequency interleaving can be applied even in the absence of time-interleaving. This is the case when the time interleaving parameter . To this end, we need a standalone feature for frequency interleaving.
Proposal 1: Define a UE feature that indicates support of Frequency Interleaving for 5G Broadcast
The feature has a granularity of “per band”, “per numerology”
The feature is optional with capability signaling.
For time interleaving, we would like to draw attention to the fact that performing time interleaving in the absence of a frequency interleaver can often be detrimental to performance (compared to not performing time-interleaving). This is so, because the design of the time-interleaver results in a scaling of the number of codeblocks by the factor , which in turn reduces the frequency span of each codeblock by a factor . This significantly impacts the frequency diversity that each codeblock can harness, if frequency interleaving is not employed. For the case when the channel is slowly varying in time (e.g., at pedestrian speeds of 3 kmph), there is limited time diversity in the system for moderate time interleaving depths, and the time interleaver cannot, in many cases, recover the loss in frequency diversity on account of scaling the TB by , as described above.
Observation 1: The TBS scaling by a factor of in the time interleaver results in each codeblock spanning a fraction in frequency, compared to a non-time-interleaved codeblock
This significantly reduces the frequency diversity that can be harnessed per codeblock, in the absence of frequency interleaving
In slowly varying channels, the gain from time diversity may not offset the above loss in frequency diversity, if the frequency interleaver is absent
In the light of this observation, we make the following proposal for the UE feature on time interleaving for 5G Broadcast.
Proposal 2: Define a UE feature that indicates support of Time Interleaving for 5G Broadcast
Support of this feature requires support of frequency interleaving as a pre-requisite
The feature has a granularity of “per band”, “per numerology”
The feature is optional with capability signaling
As agreed in RAN1, the maximum TBS supported in a TTI for MCH with time interleaving will follow that of DL-SCH. We would need to capture this (i.e., the maximum TBS supported by a UE in a TTI) as a UE feature associated with support of time interleaving.
Proposal 3: Define a UE feature that indicates the maximum TBS supported in a TTI for MCH with time interleaving
Support of this feature requires support of time interleaving as a pre-requisite
The feature is optional with capability signaling
The UE signals the maximum TBS from a set of candidate TBSs
Conclusion
In this contribution we presented our views on UE features for LTE-based 5G Broadcast Phase 2. Our proposals and observations are summarized below:
Proposal 1: Define a UE feature that indicates support of Frequency Interleaving for 5G Broadcast
The feature has a granularity of “per band”, “per numerology”
The feature is optional with capability signaling.
Observation 1: The TBS scaling by a factor of in the time interleaver results in each codeblock spanning a fraction in frequency, compared to a non-time-interleaved codeblock
This significantly reduces the frequency diversity that can be harnessed per codeblock, in the absence of frequency interleaving
In slowly varying channels, the gain from time diversity may not offset the above loss in frequency diversity, if the frequency interleaver is absent
Proposal 2: Define a UE feature that indicates support of Time Interleaving for 5G Broadcast
Support of this feature requires support of frequency interleaving as a pre-requisite
The feature has a granularity of “per band”, “per numerology”
The feature is optional with capability signaling
Proposal 3: Define a UE feature that indicates the maximum TBS supported in a TTI for MCH with time interleaving
Support of this feature requires support of time interleaving as a pre-requisite
The feature is optional with capability signaling
The UE signals the maximum TBS from a set of candidate TBSs |